Serialized Form

Package gov.sandia.cognition.algorithm

Class gov.sandia.cognition.algorithm.AbstractAnytimeAlgorithm extends AbstractIterativeAlgorithm implements Serializable

Serialized Fields

maxIterations

int maxIterations

Maximum number of iterations before stopping

Class gov.sandia.cognition.algorithm.AbstractIterativeAlgorithm extends AbstractCloneableSerializable implements Serializable

Serialized Fields

iteration

int iteration

Number of iterations the algorithm has executed.

Class gov.sandia.cognition.algorithm.AbstractParallelAlgorithm extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.algorithm.AnytimeAlgorithmWrapper extends AbstractIterativeAlgorithm implements Serializable

Serialization Methods

readResolve

protected Object readResolve()

This method is detected by the Java Serialization code and is called on deserialization. It allows the object to deal with transient values.

Serialized Fields

algorithm

AnytimeAlgorithm<ResultType> algorithm

Underlying algorithm to wrap.

Package gov.sandia.cognition.algorithm.event

Class gov.sandia.cognition.algorithm.event.AbstractIterativeAlgorithmListener extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.algorithm.event.IterationMeasurablePerformanceReporter extends AbstractIterativeAlgorithmListener implements Serializable

Serialized Fields

out

PrintStream out

The print stream to report performance to.

format

String format

The format for the performance report, passed to String.format.

Class gov.sandia.cognition.algorithm.event.IterationStartReporter extends AbstractIterativeAlgorithmListener implements Serializable

Serialized Fields

out

PrintStream out

The print stream to report performance to.

format

String format

The format for the performance report, passed to String.format.

Package gov.sandia.cognition.collection

Class gov.sandia.cognition.collection.AbstractMutableDoubleMap extends AbstractScalarMap<KeyType> implements Serializable

Serialized Fields

map

Map<K,V> map

Map backing that performs the storage.

Class gov.sandia.cognition.collection.AbstractScalarMap extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.collection.AbstractScalarMap.MapWrapper extends AbstractMap<KeyType,Double> implements Serializable

Class gov.sandia.cognition.collection.DefaultComparator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.collection.DefaultIndexer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

valueList

ArrayList<E> valueList

The list of values, which can be accessed by index.

valueToIndexMap

LinkedHashMap<K,V> valueToIndexMap

The mapping of values to their indices.

Class gov.sandia.cognition.collection.DefaultMultiCollection extends AbstractCollection<EntryType> implements Serializable

Serialized Fields

collections

List<E> collections

The set of collections that backs this collection.

Class gov.sandia.cognition.collection.DynamicArrayMap extends AbstractMap<Integer,ValueType> implements Serializable

Serialized Fields

array

Object[] array

The array underneath the mapping. Null values indicate an unassigned key.

numValues

int numValues

The number of non-null values in the array.

Class gov.sandia.cognition.collection.FiniteCapacityBuffer extends AbstractList<DataType> implements Serializable

Serialized Fields

data

Object[] data

Underlying data in the buffer

head

int head

Location of element 0

size

int size

Number of items in the list

Class gov.sandia.cognition.collection.IntegerSpan extends AbstractSet<Integer> implements Serializable

Serialized Fields

minValue

int minValue

Starting index, inclusive

maxValue

int maxValue

Stopping index, inclusive

Class gov.sandia.cognition.collection.MultiIterator extends Object implements Serializable

Serialized Fields

currentIterator

Iterator<E> currentIterator

The current iterator.

remainingIterators

LinkedList<E> remainingIterators

The iterators themselves.

Class gov.sandia.cognition.collection.NumberComparator extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.data.convert

Class gov.sandia.cognition.data.convert.AbstractDataConverter extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.data.convert.AbstractReverseCachedDataConverter extends AbstractReversibleDataConverter<InputType,OutputType> implements Serializable

Class gov.sandia.cognition.data.convert.AbstractReversibleDataConverter extends AbstractDataConverter<InputType,OutputType> implements Serializable

Class gov.sandia.cognition.data.convert.IdentityDataConverter extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.data.convert.ObjectToStringConverter extends AbstractDataConverter<Object,String> implements Serializable

Package gov.sandia.cognition.data.convert.number

Class gov.sandia.cognition.data.convert.number.DefaultBooleanToNumberConverter extends AbstractReverseCachedDataConverter<Boolean,Number,DefaultBooleanToNumberConverter.Reverse> implements Serializable

Serialized Fields

trueValue

Number trueValue

The number to use to represent a true value.

falseValue

Number falseValue

The number to use to represent a false value.

nullValue

Number nullValue

The number to use to represent a null value.

Class gov.sandia.cognition.data.convert.number.DefaultBooleanToNumberConverter.Reverse extends AbstractReversibleDataConverter<Number,Boolean> implements Serializable

Class gov.sandia.cognition.data.convert.number.StringToDoubleConverter extends AbstractReverseCachedDataConverter<String,Double,ObjectToStringConverter> implements Serializable

Class gov.sandia.cognition.data.convert.number.StringToIntegerConverter extends AbstractReverseCachedDataConverter<String,Integer,ObjectToStringConverter> implements Serializable

Package gov.sandia.cognition.data.convert.vector

Class gov.sandia.cognition.data.convert.vector.AbstractToVectorEncoder extends AbstractDataConverter<InputType,Vector> implements Serializable

Serialized Fields

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The vector factory to use to create new vectors.

Class gov.sandia.cognition.data.convert.vector.NumberConverterToVectorAdapter extends AbstractToVectorEncoder<InputType> implements Serializable

Serialized Fields

converter

DataConverter<InputType,OutputType> converter

The converter to adapt for use with Vectors.

Class gov.sandia.cognition.data.convert.vector.NumberToVectorEncoder extends AbstractToVectorEncoder<Number> implements Serializable

Class gov.sandia.cognition.data.convert.vector.UniqueBooleanVectorEncoder extends AbstractToVectorEncoder<InputType> implements Serializable

Serialized Fields

values

List<E> values

The set of possible unique values.

booleanConverter

DataToVectorEncoder<InputType> booleanConverter

The boolean encoder for the equality comparison between each of the possible values and a given input.

Package gov.sandia.cognition.evaluator

Class gov.sandia.cognition.evaluator.AbstractStatefulEvaluator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

state

CloneableSerializable state

Current state object.

Class gov.sandia.cognition.evaluator.CompositeEvaluatorList extends AbstractCloneableSerializable implements Serializable

Serialized Fields

evaluators

ArrayList<E> evaluators

The list of evaluators to compose together.

Class gov.sandia.cognition.evaluator.CompositeEvaluatorPair extends DefaultPair<Evaluator<? super InputType,? extends IntermediateType>,Evaluator<? super IntermediateType,? extends OutputType>> implements Serializable

Class gov.sandia.cognition.evaluator.CompositeEvaluatorTriple extends DefaultTriple<Evaluator<? super InputType,? extends FirstIntermediateType>,Evaluator<? super FirstIntermediateType,? extends SecondIntermediateType>,Evaluator<? super SecondIntermediateType,? extends OutputType>> implements Serializable

Class gov.sandia.cognition.evaluator.ForwardReverseEvaluatorPair extends AbstractCloneableSerializable implements Serializable

Serialized Fields

forward

Evaluator<InputType,OutputType> forward

The forward evaluator from input type to output type.

reverse

Evaluator<InputType,OutputType> reverse

The reverse evaluator from output type to input type.

Class gov.sandia.cognition.evaluator.IdentityEvaluator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.evaluator.ValueClamper extends AbstractCloneableSerializable implements Serializable

Serialized Fields

minimum

Comparable<T> minimum

The minimum possible value. May be null to not clamp to the minimum.

maximum

Comparable<T> maximum

The maximum possible value. May be null to not clamp to the maximum.

Class gov.sandia.cognition.evaluator.ValueMapper extends AbstractCloneableSerializable implements Serializable

Serialized Fields

valueMap

Map<K,V> valueMap

The map to use to map input values to output values.

Package gov.sandia.cognition.factory

Class gov.sandia.cognition.factory.DefaultFactory extends AbstractCloneableSerializable implements Serializable

Serialized Fields

createdClass

Class<T> createdClass

The class whose default constructor is used to create new objects.

Class gov.sandia.cognition.factory.PrototypeFactory extends AbstractCloneableSerializable implements Serializable

Serialized Fields

prototype

CloneableSerializable prototype

The prototype to create clones from.

Package gov.sandia.cognition.framework

Class gov.sandia.cognition.framework.AbstractCognitiveModel extends Object implements Serializable

Class gov.sandia.cognition.framework.AbstractCognitiveModelFactory extends Object implements Serializable

Serialized Fields

moduleFactories

ArrayList<E> moduleFactories

The list of CognitiveModuleFactories to use for the model.

Class gov.sandia.cognition.framework.AbstractSemanticIdentifier extends Object implements Serializable

Class gov.sandia.cognition.framework.CognitiveModelStateChangeEvent extends EventObject implements Serializable

Serialized Fields

model

CognitiveModel model

The model whose state has changed.

state

CognitiveModelState state

The new state of the model.

Class gov.sandia.cognition.framework.DefaultCogxel extends AbstractCloneableSerializable implements Serializable

Serialized Fields

semanticIdentifier

SemanticIdentifier semanticIdentifier

The SemanticIdentifier that identifies the Cogxel.

activation

double activation

The current activation of the Cogxel.

Class gov.sandia.cognition.framework.DefaultCogxelFactory extends Object implements Serializable

Class gov.sandia.cognition.framework.DefaultSemanticIdentifier extends AbstractSemanticIdentifier implements Serializable

Serialized Fields

label

SemanticLabel label

The SemanticLabel that is identified.

identifier

int identifier

The unique identifier for the semantic label.

Class gov.sandia.cognition.framework.DefaultSemanticIdentifierMap extends AbstractCloneableSerializable implements Serializable

Serialized Fields

identifierCounter

int identifierCounter

The counter of identifiers, which ensures that each one is unique.

mapping

LinkedHashMap<K,V> mapping

The mapping of semantic labels to semantic identifiers.

Class gov.sandia.cognition.framework.DefaultSemanticLabel extends Object implements Serializable

Serialized Fields

label

String label

The String label.

Class gov.sandia.cognition.framework.DefaultSemanticNetwork extends Object implements Serializable

Serialized Fields

labelNodeMap

TreeMap<K,V> labelNodeMap

The mapping of labels onto nodes in the network.

Class gov.sandia.cognition.framework.SemanticIdentifierMapEvent extends EventObject implements Serializable

Serialized Fields

map

SemanticIdentifierMap map

The map that the event happened in.

eventType

SemanticIdentifierMapEventType eventType

The type of event that this is.

identifier

SemanticIdentifier identifier

The SemanticIdentifier causing the event.

Package gov.sandia.cognition.framework.concurrent

Class gov.sandia.cognition.framework.concurrent.AbstractConcurrentCognitiveModule extends Object implements Serializable

Class gov.sandia.cognition.framework.concurrent.MultithreadedCognitiveModel extends AbstractCognitiveModelLite implements Serializable

Serialized Fields

modules

ConcurrentCognitiveModule[] modules

The CognitiveModules that are part of this model.

fixedThreadPool

ExecutorService fixedThreadPool

A thread pool for managing concurrent evaluation of modules

Class gov.sandia.cognition.framework.concurrent.MultithreadedCognitiveModelFactory extends AbstractCognitiveModelFactory implements Serializable

Serialized Fields

numThreadsInPool

int numThreadsInPool

Number of threads to use in the thread pool

Package gov.sandia.cognition.framework.learning

Class gov.sandia.cognition.framework.learning.EvaluatorBasedCognitiveModule extends AbstractConcurrentCognitiveModule implements Serializable

Serialized Fields

name

String name

Name to assign to the module

settings

EvaluatorBasedCognitiveModuleSettings<InputType,OutputType> settings

The module settings.

input

Object input

A place to temporarily store the input read in by a call to readState; this temporary store is blown away as soon as it used by evaluate, because we NEVER retain state interally across module update cycles

output

Object output

A place to temporarily store the output generated by a call to evaluate; this temporary store is blown away as soon as it used by evaluate, because we NEVER retain state interally across module update cycles

Class gov.sandia.cognition.framework.learning.EvaluatorBasedCognitiveModuleFactory extends Object implements Serializable

Serialized Fields

settings

EvaluatorBasedCognitiveModuleSettings<InputType,OutputType> settings

The settings for the module.

name

String name

Human-readable name of the module

Class gov.sandia.cognition.framework.learning.EvaluatorBasedCognitiveModuleFactoryLearner extends Object implements Serializable

Serialized Fields

learner

BatchLearner<DataType,ResultType> learner

The learner to use to learn the evaluator

inputConverter

CogxelConverter<DataType> inputConverter

The CogxelConverter used to convert from a CogxelState to InputType

outputConverter

CogxelConverter<DataType> outputConverter

The CogxelConverter used to convert OutputType to a CogxelState.

learningDataConverter

CogxelConverter<DataType> learningDataConverter

The CogxelConverter used to convert from a CogxelState to LearningDataType.

name

String name

Human-readable name for this module

Class gov.sandia.cognition.framework.learning.EvaluatorBasedCognitiveModuleSettings extends Object implements Serializable

Serialized Fields

evaluator

Evaluator<InputType,OutputType> evaluator

The evaluator to be used by the module.

inputConverter

CogxelConverter<DataType> inputConverter

The CogxelConverter used to convert from a CogxelState to InputType.

outputConverter

CogxelConverter<DataType> outputConverter

The CogxelConverter used to convert OutputType to a CogxelState.

Class gov.sandia.cognition.framework.learning.StatefulEvaluatorBasedCognitiveModule extends EvaluatorBasedCognitiveModule<InputType,OutputType> implements Serializable

Serialized Fields

stateWrapper

CognitiveModuleStateWrapper stateWrapper

A place to store the wrapper for the CognitiveModuleState that is initialized by readState and later used by writeState

Package gov.sandia.cognition.framework.learning.converter

Class gov.sandia.cognition.framework.learning.converter.AbstractCogxelConverter extends AbstractCloneableSerializable implements Serializable

Serialized Fields

semanticIdentifierMap

SemanticIdentifierMap semanticIdentifierMap

The SemanticIdentifierMap for the converter.

Class gov.sandia.cognition.framework.learning.converter.AbstractCogxelPairConverter extends AbstractCogxelConverter<PairType extends Pair<FirstType,SecondType>> implements Serializable

Serialized Fields

firstConverter

CogxelConverter<DataType> firstConverter

The CogxelConverter for the first element of the pair.

secondConverter

CogxelConverter<DataType> secondConverter

The CogxelConverter for the second element of the pair.

Class gov.sandia.cognition.framework.learning.converter.CogxelBooleanConverter extends AbstractCogxelConverter<Boolean> implements Serializable

Serialized Fields

label

SemanticLabel label

The label of the Cogxel to convert.

identifier

SemanticIdentifier identifier

The semantic identifier of the Cogxel to convert.

cogxelFactory

CogxelFactory cogxelFactory

The CogxelFactory to use.

Class gov.sandia.cognition.framework.learning.converter.CogxelDoubleConverter extends Object implements Serializable

Serialized Fields

label

SemanticLabel label

The label of the Cogxel to convert.

identifier

SemanticIdentifier identifier

The semantic identifier of the Cogxel to convert.

semanticIdentifierMap

SemanticIdentifierMap semanticIdentifierMap

The SemanticIdentifierMap for the converter.

cogxelFactory

CogxelFactory cogxelFactory

The CogxelFactory to use.

Class gov.sandia.cognition.framework.learning.converter.CogxelInputOutputPairConverter extends AbstractCogxelPairConverter<InputType,OutputType,InputOutputPair<InputType,OutputType>> implements Serializable

Class gov.sandia.cognition.framework.learning.converter.CogxelMatrixConverter extends Object implements Serializable

Serialized Fields

columnConverters

ArrayList<E> columnConverters

Collection CogxelVectorConverters that convert the columns of the matrix

semanticIdentifierMap

SemanticIdentifierMap semanticIdentifierMap

SemanticIdentifierMap for the converter

Class gov.sandia.cognition.framework.learning.converter.CogxelTargetEstimatePairConverter extends AbstractCogxelPairConverter<TargetType,EstimateType,TargetEstimatePair<TargetType,EstimateType>> implements Serializable

Class gov.sandia.cognition.framework.learning.converter.CogxelVectorCollectionConverter extends Object implements Serializable

Serialized Fields

cogxelVectorConverters

Collection<E> cogxelVectorConverters

Collection of CogxelVectorConverters that do the heavy lifting

Class gov.sandia.cognition.framework.learning.converter.CogxelVectorConverter extends DefaultVectorFactoryContainer implements Serializable

Serialized Fields

labels

ArrayList<E> labels

The labels for each of the elements in the vector.

semanticIdentifierMap

SemanticIdentifierMap semanticIdentifierMap

The current SemanticIdentifierMap.

identifiers

ArrayList<E> identifiers

The list of SemanticIdentifiers, whose positions correspond to positions in the vector.

identifierToIndexMap

HashMap<K,V> identifierToIndexMap

The mapping of SemanticIdentifiers to vector indices.

cogxelFactory

CogxelFactory cogxelFactory

The CogxelFactory to use.

Class gov.sandia.cognition.framework.learning.converter.CogxelWeightedInputOutputPairConverter extends Object implements Serializable

Serialized Fields

pairConverter

CogxelInputOutputPairConverter<InputType,OutputType> pairConverter

Creates the underlying InputOutputPair

weightConverter

CogxelConverter<DataType> weightConverter

Adds the weight to the InputOutputPair

Package gov.sandia.cognition.framework.lite

Class gov.sandia.cognition.framework.lite.AbstractCognitiveModelLite extends AbstractCognitiveModel implements Serializable

Serialized Fields

numModules

int numModules

The number of modules.

semanticIdentifierMap

DefaultSemanticIdentifierMap semanticIdentifierMap

The semantic identifier database.

state

CognitiveModelLiteState state

The current state of the model.

Class gov.sandia.cognition.framework.lite.AbstractSemanticMemoryLite extends Object implements Serializable

Serialized Fields

semanticIdentifierMap

SemanticIdentifierMap semanticIdentifierMap

The mapping of semantic labels to semantic identifiers.

recognizer

PatternRecognizerLite recognizer

The PatternRecognizerLite that is used.

outputIdentifiers

ArrayList<E> outputIdentifiers

The mapping of output vector entries to identifiers.

Class gov.sandia.cognition.framework.lite.ArrayBasedCognitiveModelInput extends Object implements Serializable

Serialized Fields

identifiers

SemanticIdentifier[] identifiers

The identifiers to use as inputs.

values

double[] values

The activation values of the inputs.

Class gov.sandia.cognition.framework.lite.ArrayBasedPerceptionModule extends AbstractConcurrentCognitiveModule implements Serializable

Serialized Fields

semanticIdentifierMap

SemanticIdentifierMap semanticIdentifierMap

The semantic identifier map. Currently unused, but provided because there may be a need to support SemanticLabels instead of identifiers later on.

cogxelFactory

CogxelFactory cogxelFactory

The factory for creating Cogxels for perception.

input

ArrayBasedCognitiveModelInput input

A place to temporarily store the input read in by a call to readState; this temporary store is blown away as soon as it used by writeState, because we NEVER retain state interally across module update cycles

Class gov.sandia.cognition.framework.lite.ArrayBasedPerceptionModuleFactory extends Object implements Serializable

Serialized Fields

cogxelFactory

CogxelFactory cogxelFactory

The CogxelFactory for the module to use in creating Cogxels.

Class gov.sandia.cognition.framework.lite.BooleanActivatableCogxel extends DefaultCogxel implements Serializable

Serialized Fields

activated

boolean activated

A boolean indicating whether the cogxel is activated.

Class gov.sandia.cognition.framework.lite.BooleanActivatableCogxelFactory extends Object implements Serializable

Class gov.sandia.cognition.framework.lite.CognitiveModelLite extends AbstractCognitiveModelLite implements Serializable

Serialized Fields

modules

CognitiveModule[] modules

The CognitiveModules that are part of this model.

Class gov.sandia.cognition.framework.lite.CognitiveModelLiteFactory extends AbstractCognitiveModelFactory implements Serializable

Class gov.sandia.cognition.framework.lite.CognitiveModelLiteState extends AbstractCloneableSerializable implements Serializable

Serialized Fields

initialized

boolean initialized

A flag indicating if the state has been initialized or not.

input

CognitiveModelInput input

The input to the model.

cogxels

CogxelStateLite cogxels

The state of the Cogxels.

moduleStatesArray

CognitiveModuleState[] moduleStatesArray

The states of the modules.

Class gov.sandia.cognition.framework.lite.CognitiveModuleStateWrapper extends AbstractCloneableSerializable implements Serializable

Serialized Fields

internalState

CloneableSerializable internalState

The internal state object that is wrapped.

Class gov.sandia.cognition.framework.lite.CogxelStateLite extends AbstractCloneableSerializable implements Serializable

Serialized Fields

cogxels

DynamicArrayMap<ValueType> cogxels

The underlying mapping of identifiers to cogxels.

Class gov.sandia.cognition.framework.lite.MutableSemanticMemoryLite extends AbstractSemanticMemoryLite implements Serializable

Serialized Fields

mutableRecognizer

MutablePatternRecognizerLite mutableRecognizer

The mutable pattern recognizer to modify.

identifierToInputIndexMap

HashMap<K,V> identifierToInputIndexMap

The mapping of identifier number to input vector index number.

nodeSet

HashSet<E> nodeSet

The set of nodes used by the memory.

Class gov.sandia.cognition.framework.lite.MutableSemanticMemoryLiteFactory extends Object implements Serializable

Serialized Fields

recognizer

MutablePatternRecognizerLite recognizer

The module settings.

Class gov.sandia.cognition.framework.lite.SharedSemanticMemoryLite extends AbstractSemanticMemoryLite implements Serializable

Serialized Fields

sharedSettings

SharedSemanticMemoryLiteSettings sharedSettings

The shared settings for this semantic memory.

minIdentifier

int minIdentifier

The minimum identifier number.

maxIdentifier

int maxIdentifier

The maximum identifier number.

identifierToInputIndexMap

int[] identifierToInputIndexMap

The mapping of input indentifier to input index.

Class gov.sandia.cognition.framework.lite.SharedSemanticMemoryLiteFactory extends AbstractCloneableSerializable implements Serializable

Serialized Fields

settings

SharedSemanticMemoryLiteSettings settings

The shared settings between the modules.

Class gov.sandia.cognition.framework.lite.SharedSemanticMemoryLiteSettings extends AbstractCloneableSerializable implements Serializable

Serialized Fields

recognizer

PatternRecognizerLite recognizer

The recognizer to use in the memory.

Class gov.sandia.cognition.framework.lite.SimplePatternRecognizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

nodes

ArrayList<E> nodes

The nodes in the recognition network, sorted by how they are to appear in the vector.

nodeToIDMap

HashMap<K,V> nodeToIDMap

The mapping of semantic labels to their index in the vector.

matrix

Matrix matrix

The matrix underlying the pattern recognizer.

nodesChangedSinceLastUpdate

boolean nodesChangedSinceLastUpdate

True if the set of nodes has changed since the last update.

Class gov.sandia.cognition.framework.lite.SimplePatternRecognizerState extends AbstractCloneableSerializable implements Serializable

Serialized Fields

labels

ArrayList<E> labels

The array of semantic labels that the state vector has. This is stored as a private copy in the state object because the module can have labels added or removed. If that happens then the module will need to update the state based on the new labels.

stateVector

Vector stateVector

The state vector.

Class gov.sandia.cognition.framework.lite.VectorBasedCognitiveModelInput extends AbstractCloneableSerializable implements Serializable

Serialized Fields

identifiers

SemanticIdentifier[] identifiers

Identifiers to use associate with the inputs

values

Vector values

Activation values of the inputs

Class gov.sandia.cognition.framework.lite.VectorBasedPerceptionModule extends Object implements Serializable

Serialized Fields

cogxelFactory

CogxelFactory cogxelFactory

Factory for creating Cogxels for perception

Class gov.sandia.cognition.framework.lite.VectorBasedPerceptionModuleFactory extends Object implements Serializable

Serialized Fields

cogxelFactory

CogxelFactory cogxelFactory

CogxelFactory for the module to use in creating Cogxels

Package gov.sandia.cognition.io

Class gov.sandia.cognition.io.CSVParseException extends Exception implements Serializable

Package gov.sandia.cognition.io.serialization

Class gov.sandia.cognition.io.serialization.AbstractFileSerializationHandler extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.io.serialization.AbstractStreamSerializationHandler extends AbstractFileSerializationHandler<SerializedType> implements Serializable

Class gov.sandia.cognition.io.serialization.AbstractTextSerializationHandler extends AbstractStreamSerializationHandler<SerializedType> implements Serializable

Serialization Methods

readObject

public Object readObject(InputStream stream)
                  throws IOException,
                         ClassNotFoundException

Throws:

IOException - If there is an i/o error.

ClassNotFoundException - If a class cannot be found.

writeObject

public void writeObject(OutputStream stream,
                        SerializedType object)
                 throws IOException

Throws:

IOException - If there is an i/o error.

Class gov.sandia.cognition.io.serialization.GZIPSerializationHandler extends AbstractStreamSerializationHandler<SerializedType> implements Serializable

Serialization Methods

readObject

public Object readObject(InputStream stream)
                  throws IOException,
                         ClassNotFoundException

Throws:

IOException - If there is an i/o error.

ClassNotFoundException - If a class cannot be found.

writeObject

public void writeObject(OutputStream stream,
                        SerializedType object)
                 throws IOException

Throws:

IOException - If there is an i/o error.

Serialized Fields

baseHandler

StreamSerializationHandler<SerializedType> baseHandler

The base handler that is being wrapped in a GZip.

Class gov.sandia.cognition.io.serialization.JavaDefaultBinarySerializationHandler extends AbstractStreamSerializationHandler<Serializable> implements Serializable

Serialization Methods

readObject

public Object readObject(InputStream stream)
                  throws IOException,
                         ClassNotFoundException

Throws:

IOException - If there is an i/o error.

ClassNotFoundException - If a class cannot be found.

writeObject

public void writeObject(OutputStream stream,
                        Serializable object)
                 throws IOException

Throws:

IOException - If there is an i/o error.

Class gov.sandia.cognition.io.serialization.XStreamSerializationHandler extends AbstractTextSerializationHandler<Serializable> implements Serializable

Serialization Methods

readObject

public Object readObject(Reader reader)
                  throws IOException

Throws:

IOException - If there is an i/o error.

writeObject

public void writeObject(Writer writer,
                        Serializable object)
                 throws IOException

Throws:

IOException - If there is an i/o error.

Package gov.sandia.cognition.learning.algorithm

Class gov.sandia.cognition.learning.algorithm.AbstractAnytimeBatchLearner extends AbstractAnytimeAlgorithm<ResultType> implements Serializable

Serialized Fields

keepGoing

boolean keepGoing

Indicates whether or not the learner should make another step.

data

Object data

The data to learn from.

Class gov.sandia.cognition.learning.algorithm.AbstractAnytimeSupervisedBatchLearner extends AbstractAnytimeBatchLearner<Collection<? extends InputOutputPair<? extends InputType,OutputType>>,ResultType extends Evaluator<? super InputType,? extends OutputType>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.AbstractBatchAndIncrementalLearner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.AbstractBatchLearnerContainer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

learner

BatchLearner<DataType,ResultType> learner

The wrapped learner.

Class gov.sandia.cognition.learning.algorithm.AbstractSupervisedBatchAndIncrementalLearner extends AbstractBatchAndIncrementalLearner<InputOutputPair<? extends InputType,OutputType>,ResultType extends Evaluator<? super InputType,? extends OutputType>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.CompositeBatchLearnerPair extends AbstractCloneableSerializable implements Serializable

Serialized Fields

firstLearner

BatchLearner<DataType,ResultType> firstLearner

The first learner that is trained on the input data.

secondLearner

BatchLearner<DataType,ResultType> secondLearner

The second learner that is trained on the output of the evaluator created by the first learner.

Class gov.sandia.cognition.learning.algorithm.InputOutputTransformedBatchLearner extends AbstractBatchLearnerContainer<BatchLearner<? super Collection<? extends InputOutputPair<? extends TransformedInputType,TransformedOutputType>>,? extends Evaluator<? super TransformedInputType,? extends TransformedOutputType>>> implements Serializable

Serialized Fields

inputLearner

BatchLearner<DataType,ResultType> inputLearner

The unsupervised learning algorithm for creating the input transformation.

outputLearner

BatchLearner<DataType,ResultType> outputLearner

The unsupervised learning algorithm for creating the output transformation, which must be reversible for evaluation.

Class gov.sandia.cognition.learning.algorithm.SequencePredictionLearner extends AbstractBatchLearnerContainer<BatchLearner<? super Collection<? extends InputOutputPair<? extends DataType,DataType>>,? extends LearnedType>> implements Serializable

Serialized Fields

predictionHorizon

int predictionHorizon

The prediction horizon, which is the number of samples in the future to try to learn to predict. It must be a positive number.

Class gov.sandia.cognition.learning.algorithm.TimeSeriesPredictionLearner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

predictionHorizon

int predictionHorizon

Number of samples into the future to predict.

supervisedLearner

SupervisedBatchLearner<InputType,OutputType,ResultType extends Evaluator<? super InputType,? extends OutputType>> supervisedLearner

Learning algorithm that does the heavy lifting.

Package gov.sandia.cognition.learning.algorithm.annealing

Class gov.sandia.cognition.learning.algorithm.annealing.SimulatedAnnealer extends AbstractAnytimeBatchLearner<CostParametersType,AnnealedType> implements Serializable

Serialized Fields

cost

CostFunction<EvaluatedType,CostParametersType> cost

The cost function to minimize.

perturber

Perturber<PerturbedType> perturber

The perturbing function to use to perturb the objects.

temperature

double temperature

The current temperature.

maxIterationsWithoutImprovement

int maxIterationsWithoutImprovement

The maximum number of iterations to go without improvement before stopping.

iterationsWithoutImprovement

int iterationsWithoutImprovement

The number of iterations since the last improvement.

coolingFactor

double coolingFactor

The cooling factor applied at each step.

random

Random random

The random number generator to use.

bestSoFar

Object bestSoFar

The best state found so far.

bestSoFarScore

double bestSoFarScore

The score for the best state found so far.

current

Object current

The current state.

currentScore

double currentScore

The score of the current state.

Class gov.sandia.cognition.learning.algorithm.annealing.VectorizablePerturber extends AbstractRandomized implements Serializable

Serialized Fields

covarianceSqrt

Matrix covarianceSqrt

The covariance square root matrix to use while perturbing.

Package gov.sandia.cognition.learning.algorithm.baseline

Class gov.sandia.cognition.learning.algorithm.baseline.ConstantLearner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

value

Object value

The result of learning.

Class gov.sandia.cognition.learning.algorithm.baseline.IdentityLearner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.baseline.MeanLearner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.baseline.MostFrequentLearner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.baseline.WeightedMeanLearner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.baseline.WeightedMostFrequentLearner extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.algorithm.bayes

Class gov.sandia.cognition.learning.algorithm.bayes.DiscreteNaiveBayesCategorizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

conditionalProbabilities

Map<K,V> conditionalProbabilities

Class conditional probability table.

priorProbabilities

DefaultDataDistribution<KeyType> priorProbabilities

Table of category priors.

inputDimensionality

int inputDimensionality

Assumed dimensionality of the inputs.

Class gov.sandia.cognition.learning.algorithm.bayes.DiscreteNaiveBayesCategorizer.Learner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.bayes.VectorNaiveBayesCategorizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

priors

DataDistribution<DataType> priors

The prior distribution for the categorizer.

conditionals

Map<K,V> conditionals

The mapping of category to the conditional distribution for the category with one probability density function for each dimension.

Class gov.sandia.cognition.learning.algorithm.bayes.VectorNaiveBayesCategorizer.BatchGaussianLearner extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.bayes.VectorNaiveBayesCategorizer.Learner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

distributionEstimator

DistributionEstimator<ObservationType,DistributionType extends Distribution<? extends ObservationType>> distributionEstimator

The distributionLearner for the distribution of each dimension of each category.

Class gov.sandia.cognition.learning.algorithm.bayes.VectorNaiveBayesCategorizer.OnlineLearner extends AbstractBatchAndIncrementalLearner<InputOutputPair<? extends Vectorizable,CategoryType>,VectorNaiveBayesCategorizer<CategoryType,DistributionType extends UnivariateProbabilityDensityFunction>> implements Serializable

Serialized Fields

distributionLearner

IncrementalLearner<DataType,ResultType> distributionLearner

The incremental learner for the distribution used to represent each dimension. By the generic, it must learn a univariate probability density function.

Package gov.sandia.cognition.learning.algorithm.clustering

Class gov.sandia.cognition.learning.algorithm.clustering.AffinityPropagation extends AbstractAnytimeBatchLearner<Collection<? extends DataType>,Collection<CentroidCluster<DataType>>> implements Serializable

Serialized Fields

divergence

DivergenceFunction<FirstType,SecondType> divergence

The divergence function to use.

selfDivergence

double selfDivergence

The value to use for self-divergence. Controls the number of clusters created.

dampingFactor

double dampingFactor

The damping factor (lambda). It must be between 0.0 and 1.0.

oneMinusDampingFactor

double oneMinusDampingFactor

The cached value of one minus the damping factor.

examples

ArrayList<E> examples

The examples.

similarities

double[][] similarities

The array of example-example similarities.

responsibilities

double[][] responsibilities

The array of example-example responsibilities.

availabilities

double[][] availabilities

The array of example-example availabilities.

assignments

int[] assignments

The assignments of each example to an exemplar (cluster).

changedCount

int changedCount

The number of examples that have changed assignments in the last iteration.

clusters

HashMap<K,V> clusters

The clusters that have been found so far. It is a sparse mapping since we expect there to be few clusters.

Class gov.sandia.cognition.learning.algorithm.clustering.AgglomerativeClusterer extends AbstractAnytimeBatchLearner<Collection<? extends DataType>,Collection<ClusterType extends Cluster<DataType>>> implements Serializable

Serialized Fields

divergenceFunction

ClusterToClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> divergenceFunction

The divergence function used to find the distance between two clusters.

creator

ClusterCreator<ClusterType extends Cluster<DataType>,DataType> creator

The merger used to merge two clusters into one element.

minNumClusters

int minNumClusters

The minimum number of clusters allowed.

maxMinDistance

double maxMinDistance

The maximum minimum distance between clusters allowed.

clusters

ArrayList<E> clusters

The current set of clusters.

clustersHierarchy

ArrayList<E> clustersHierarchy

The current set of hierarchical clusters.

Class gov.sandia.cognition.learning.algorithm.clustering.AgglomerativeClusterer.HierarchyNode extends BinaryClusterHierarchyNode<DataType,ClusterType extends Cluster<DataType>> implements Serializable

Serialized Fields

childrenDivergence

double childrenDivergence

The divergence between the two children, if they exist.

Class gov.sandia.cognition.learning.algorithm.clustering.DirichletProcessClustering extends AnytimeAlgorithmWrapper<Collection<GaussianCluster>,DirichletProcessMixtureModel<Vector>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.KMeansClusterer extends AbstractAnytimeBatchLearner<Collection<? extends DataType>,Collection<ClusterType extends Cluster<DataType>>> implements Serializable

Serialized Fields

numRequestedClusters

int numRequestedClusters

The number of clusters requested.

initializer

FixedClusterInitializer<ClusterType extends Cluster<DataType>,DataType> initializer

The initializer for the algorithm.

divergenceFunction

ClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> divergenceFunction

The divergence function between cluster being used.

creator

ClusterCreator<ClusterType extends Cluster<DataType>,DataType> creator

The cluster creator for creating clusters.

clusters

ArrayList<E> clusters

The current set of clusters.

assignments

int[] assignments

The current assignments of elements to clusters.

clusterCounts

int[] clusterCounts

The current number of elements assigned to each cluster.

numChanged

int numChanged

Returns the number of samples that changed assignment between iterations

Class gov.sandia.cognition.learning.algorithm.clustering.KMeansClustererWithRemoval extends KMeansClusterer<DataType,ClusterType extends Cluster<DataType>> implements Serializable

Serialized Fields

removalThreshold

double removalThreshold

fraction of the expected number of data points assigned to a cluster below which the cluster will be removed. (Suppose there are 1000 datapoint, 10 clusters, and removalThreshold=0.1. A cluster may be removed only if is has membership less than 0.1*1000/10= 10 elements assigned to it.)

Class gov.sandia.cognition.learning.algorithm.clustering.KMeansFactory extends AbstractRandomized implements Serializable

Serialized Fields

numClusters

int numClusters

Number of clusters to use. Must be positive.

Class gov.sandia.cognition.learning.algorithm.clustering.OptimizedKMeansClusterer extends KMeansClusterer<DataType,CentroidCluster<DataType>> implements Serializable

Serialized Fields

metric

Metric<EvaluatedType> metric

The metric being used.

lowerBounds

double[][] lowerBounds

The lower bounds on the distances to the clusters.

upperBounds

double[] upperBounds

The upper bounds on the distance to the current assigned cluster.

clusterDistances

double[][] clusterDistances

The distances between clusters.

Class gov.sandia.cognition.learning.algorithm.clustering.ParallelizedKMeansClusterer extends KMeansClusterer<DataType,ClusterType extends Cluster<DataType>> implements Serializable

Serialized Fields

assignmentTasks

ArrayList<E> assignmentTasks

Parallel tasks that assign the data points to clusters

clusterCreatorTask

ArrayList<E> clusterCreatorTask

Parallel tasks that creates clusters from the assigned data points

assignmentList

Collection<E> assignmentList

ArrayList of assignments from the subtasks

newAssignments

int[] newAssignments

Array of new assignments

Class gov.sandia.cognition.learning.algorithm.clustering.PartitionalClusterer extends AbstractAnytimeBatchLearner<Collection<? extends DataType>,Collection<ClusterType extends Cluster<DataType>>> implements Serializable

Serialized Fields

divergenceFunction

ClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> divergenceFunction

The divergence function used to find the distance between two clusters.

creator

IncrementalClusterCreator<ClusterType extends Cluster<DataType>,DataType> creator

The merger used to merge two clusters into one element.

minClusterSize

int minClusterSize

The minimum number of elements per cluster allowed.

maxCriterionDecrease

double maxCriterionDecrease

The maximum decrease in training criterion allowed.

random

Random random

The source of randomness used during initial partitioning.

Package gov.sandia.cognition.learning.algorithm.clustering.cluster

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.CentroidCluster extends DefaultCluster<ClusterType> implements Serializable

Serialized Fields

centroid

Object centroid

The center of the cluster.

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.DefaultCluster extends AbstractCloneableSerializable implements Serializable

Serialized Fields

index

int index

The index of the cluster in the collection of clusters.

members

ArrayList<E> members

The members of the cluster.

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.DefaultClusterCreator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.DefaultIncrementalClusterCreator extends DefaultClusterCreator<DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.GaussianCluster extends DefaultCluster<Vector> implements Serializable

Serialized Fields

gaussian

MultivariateGaussian.PDF gaussian

The Gaussian distribution that the cluster represents.

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.GaussianClusterCreator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

defaultCovariance

double defaultCovariance

amount to add to the diagonals of the covariance matrix, typically on the order of 1e-4.

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.MedoidClusterCreator extends DefaultDivergenceFunctionContainer<DataType,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.cluster.VectorMeanCentroidClusterCreator extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.algorithm.clustering.divergence

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.AbstractClusterToClusterDivergenceFunction extends DefaultDivergenceFunctionContainer<DataType,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.CentroidClusterDivergenceFunction extends DefaultDivergenceFunctionContainer<DataType,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.ClusterCentroidDivergenceFunction extends AbstractClusterToClusterDivergenceFunction<CentroidCluster<DataType>,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.ClusterCompleteLinkDivergenceFunction extends AbstractClusterToClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.ClusterMeanLinkDivergenceFunction extends AbstractClusterToClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.ClusterSingleLinkDivergenceFunction extends AbstractClusterToClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.divergence.GaussianClusterDivergenceFunction extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.algorithm.clustering.hierarchy

Class gov.sandia.cognition.learning.algorithm.clustering.hierarchy.AbstractClusterHierarchyNode extends AbstractCloneableSerializable implements Serializable

Serialized Fields

cluster

Cluster<ClusterType> cluster

The cluster associated with the node.

Class gov.sandia.cognition.learning.algorithm.clustering.hierarchy.BinaryClusterHierarchyNode extends AbstractClusterHierarchyNode<DataType,ClusterType extends Cluster<DataType>> implements Serializable

Serialized Fields

firstChild

ClusterHierarchyNode<DataType,ClusterType extends Cluster<DataType>> firstChild

The first child node.

secondChild

ClusterHierarchyNode<DataType,ClusterType extends Cluster<DataType>> secondChild

The second child node.

Class gov.sandia.cognition.learning.algorithm.clustering.hierarchy.DefaultClusterHierarchyNode extends AbstractClusterHierarchyNode<DataType,ClusterType extends Cluster<DataType>> implements Serializable

Serialized Fields

children

List<E> children

The list of children.

Package gov.sandia.cognition.learning.algorithm.clustering.initializer

Class gov.sandia.cognition.learning.algorithm.clustering.initializer.AbstractMinDistanceFixedClusterInitializer extends DefaultDivergenceFunctionContainer<DataType,DataType> implements Serializable

Serialized Fields

creator

ClusterCreator<ClusterType extends Cluster<DataType>,DataType> creator

The ClusterCreator to create the initial clusters from.

random

Random random

The random number generator to use.

Class gov.sandia.cognition.learning.algorithm.clustering.initializer.DistanceSamplingClusterInitializer extends AbstractMinDistanceFixedClusterInitializer<ClusterType extends Cluster<DataType>,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.initializer.GreedyClusterInitializer extends AbstractMinDistanceFixedClusterInitializer<ClusterType extends Cluster<DataType>,DataType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.clustering.initializer.NeighborhoodGaussianClusterInitializer extends AbstractRandomized implements Serializable

Serialized Fields

defaultCovariance

double defaultCovariance

default diagonal covariance scaling factor

randomRange

double randomRange

range of the neighborhood from which to place the cluster

Package gov.sandia.cognition.learning.algorithm.confidence

Class gov.sandia.cognition.learning.algorithm.confidence.AdaptiveRegularizationOfWeights extends AbstractSupervisedBatchAndIncrementalLearner<Vectorizable,Boolean,DefaultConfidenceWeightedBinaryCategorizer> implements Serializable

Serialized Fields

r

double r

The r parameter that controls regularization weight. Must be positive.

Class gov.sandia.cognition.learning.algorithm.confidence.ConfidenceWeightedDiagonalDeviation extends AbstractSupervisedBatchAndIncrementalLearner<Vectorizable,Boolean,DiagonalConfidenceWeightedBinaryCategorizer> implements Serializable

Serialized Fields

confidence

double confidence

The confidence to use for updating. Must be in [0.5, 1]. Called eta in the paper.

defaultVariance

double defaultVariance

The default variance, which the diagonal of the covariance matrix is initialized to. Must be positive. Called a in the paper.

phi

double phi

Phi is the standard score computed from the confidence.

psi

double psi

Psi is the cached value 1 + phi^2 / 2.

epsilon

double epsilon

Epsilon is the cached value 1 + phi^2.

Class gov.sandia.cognition.learning.algorithm.confidence.ConfidenceWeightedDiagonalDeviationProject extends ConfidenceWeightedDiagonalDeviation implements Serializable

Class gov.sandia.cognition.learning.algorithm.confidence.ConfidenceWeightedDiagonalVariance extends AbstractSupervisedBatchAndIncrementalLearner<Vectorizable,Boolean,DiagonalConfidenceWeightedBinaryCategorizer> implements Serializable

Serialized Fields

confidence

double confidence

The confidence to use for updating. Must be in [0, 1]. Called eta in the paper.

defaultVariance

double defaultVariance

The default variance, which the diagonal of the covariance matrix is initialized to. Must be positive. Called a in the paper.

phi

double phi

Phi is the standard score computed from the confidence.

Class gov.sandia.cognition.learning.algorithm.confidence.ConfidenceWeightedDiagonalVarianceProject extends ConfidenceWeightedDiagonalVariance implements Serializable

Package gov.sandia.cognition.learning.algorithm.ensemble

Class gov.sandia.cognition.learning.algorithm.ensemble.AbstractBaggingLearner extends AbstractAnytimeSupervisedBatchLearner<InputType,OutputType,EnsembleType extends Evaluator<? super InputType,? extends OutputType>> implements Serializable

Serialized Fields

learner

BatchLearner<DataType,ResultType> learner

The learner to use to create the categorizer for each iteration.

percentToSample

double percentToSample

The percentage of the data to sample with replacement on each iteration. Must be positive. Represented as a floating point number with 1.0 meaning 100%.

random

Random random

The random number generator to use.

Class gov.sandia.cognition.learning.algorithm.ensemble.AbstractUnweightedEnsemble extends AbstractCloneableSerializable implements Serializable

Serialized Fields

members

List<E> members

The members of the ensemble.

Class gov.sandia.cognition.learning.algorithm.ensemble.AbstractWeightedEnsemble extends AbstractCloneableSerializable implements Serializable

Serialized Fields

members

List<E> members

The members of the ensemble.

Class gov.sandia.cognition.learning.algorithm.ensemble.AdaBoost extends AbstractAnytimeSupervisedBatchLearner<InputType,Boolean,WeightedBinaryEnsemble<InputType,Evaluator<? super InputType,? extends Boolean>>> implements Serializable

Serialized Fields

weakLearner

BatchLearner<DataType,ResultType> weakLearner

The "weak learner" that must learn from the weighted input-output pairs on each iteration.

Class gov.sandia.cognition.learning.algorithm.ensemble.AdditiveEnsemble extends AbstractUnweightedEnsemble<MemberType extends Evaluator<? super InputType,? extends Number>> implements Serializable

Serialized Fields

bias

double bias

The initial offset value that the ensemble outputs are added to.

Class gov.sandia.cognition.learning.algorithm.ensemble.AveragingEnsemble extends AbstractUnweightedEnsemble<MemberType extends Evaluator<? super InputType,? extends Number>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.ensemble.BaggingCategorizerLearner extends AbstractBaggingLearner<InputType,CategoryType,Evaluator<? super InputType,? extends CategoryType>,WeightedVotingCategorizerEnsemble<InputType,CategoryType,Evaluator<? super InputType,? extends CategoryType>>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.ensemble.BaggingRegressionLearner extends AbstractBaggingLearner<InputType,Double,Evaluator<? super InputType,? extends Number>,AveragingEnsemble<InputType,Evaluator<? super InputType,? extends Number>>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.ensemble.BinaryBaggingLearner extends AbstractBaggingLearner<InputType,Boolean,Evaluator<? super InputType,? extends Boolean>,WeightedBinaryEnsemble<InputType,Evaluator<? super InputType,? extends Boolean>>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.ensemble.BinaryCategorizerSelector extends AbstractCloneableSerializable implements Serializable

Serialized Fields

categorizers

Collection<E> categorizers

The collection of categorizers to evaluate and select from.

Class gov.sandia.cognition.learning.algorithm.ensemble.CategoryBalancedBaggingLearner extends BaggingCategorizerLearner<InputType,CategoryType> implements Serializable

Serialized Fields

categoryList

ArrayList<E> categoryList

The list of categories.

dataPerCategory

HashMap<K,V> dataPerCategory

The mapping of categories to indices of examples belonging to the category.

Class gov.sandia.cognition.learning.algorithm.ensemble.CategoryBalancedIVotingLearner extends IVotingCategorizerLearner<InputType,CategoryType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.ensemble.IVotingCategorizerLearner extends AbstractAnytimeSupervisedBatchLearner<InputType,CategoryType,WeightedVotingCategorizerEnsemble<InputType,CategoryType,Evaluator<? super InputType,? extends CategoryType>>> implements Serializable

Serialized Fields

learner

BatchLearner<DataType,ResultType> learner

The learner used to produce each ensemble member.

percentToSample

double percentToSample

The percent to sample on each iteration.

proportionIncorrectInSample

double proportionIncorrectInSample

The proportion of incorrect examples in each sample. Must be between 0.0 and 1.0 (inclusive).

voteOutOfBagOnly

boolean voteOutOfBagOnly

Controls whether or not an ensemble member can vote on items it was trained on during learning. By default, the ensemble member can only vote on out-of-bag values.

counterFactory

Factory<CreatedType> counterFactory

Factory for counting votes.

random

Random random

The random number generator to use.

Class gov.sandia.cognition.learning.algorithm.ensemble.IVotingCategorizerLearner.OutOfBagErrorStoppingCriteria extends AbstractIterativeAlgorithmListener implements Serializable

Serialized Fields

smoothingWindowSize

int smoothingWindowSize

The size of window of data to look at to determine if learning has hit a minimum.

Class gov.sandia.cognition.learning.algorithm.ensemble.MultiCategoryAdaBoost extends AbstractAnytimeSupervisedBatchLearner<InputType,CategoryType,WeightedVotingCategorizerEnsemble<InputType,CategoryType,Evaluator<? super InputType,? extends CategoryType>>> implements Serializable

Serialized Fields

weakLearner

BatchLearner<DataType,ResultType> weakLearner

The "weak learner" that must learn from the weighted input-output pairs on each iteration.

Class gov.sandia.cognition.learning.algorithm.ensemble.OnlineBaggingCategorizerLearner extends AbstractSupervisedBatchAndIncrementalLearner<InputType,CategoryType,VotingCategorizerEnsemble<InputType,CategoryType,MemberType extends Evaluator<? super InputType,? extends CategoryType>>> implements Serializable

Serialized Fields

learner

IncrementalLearner<DataType,ResultType> learner

The base learner used for each ensemble member.

ensembleSize

int ensembleSize

The size of the ensemble to create. Must be positive.

percentToSample

double percentToSample

The percentage of the data to sample for each ensemble member. Must be positive. Used as a parameter to the Poisson distribution to determine the number of samples for each ensemble member.

random

Random random

The random number generator to use.

Class gov.sandia.cognition.learning.algorithm.ensemble.VotingCategorizerEnsemble extends AbstractDiscriminantCategorizer<InputType,CategoryType,Double> implements Serializable

Serialized Fields

members

List<E> members

The members of the ensemble.

Class gov.sandia.cognition.learning.algorithm.ensemble.WeightedAdditiveEnsemble extends AbstractWeightedEnsemble<MemberType extends Evaluator<? super InputType,? extends Number>> implements Serializable

Serialized Fields

bias

double bias

The initial offset value that the ensemble outputs are added to.

Class gov.sandia.cognition.learning.algorithm.ensemble.WeightedAveragingEnsemble extends AbstractWeightedEnsemble<MemberType extends Evaluator<? super InputType,? extends Number>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.ensemble.WeightedBinaryEnsemble extends AbstractDiscriminantBinaryCategorizer<InputType> implements Serializable

Serialized Fields

members

List<E> members

The members of the ensemble.

Class gov.sandia.cognition.learning.algorithm.ensemble.WeightedVotingCategorizerEnsemble extends AbstractCategorizer<InputType,CategoryType> implements Serializable

Serialized Fields

members

List<E> members

The members of the ensemble.

Package gov.sandia.cognition.learning.algorithm.genetic

Class gov.sandia.cognition.learning.algorithm.genetic.EvaluatedGenome extends Object implements Serializable

Serialized Fields

genome

Object genome

The genome that was evaluated.

cost

double cost

The cost associated with a Genome.

Class gov.sandia.cognition.learning.algorithm.genetic.GeneticAlgorithm extends AbstractAnytimeBatchLearner<CostParametersType,GenomeType> implements Serializable

Serialized Fields

costFunction

CostFunction<EvaluatedType,CostParametersType> costFunction

The cost function for genomes.

reproducer

Reproducer<GenomeType> reproducer

The reproduction function for genomes.

bestSoFar

EvaluatedGenome<GenomeType> bestSoFar

The best genome found so far.

maxIterationsWithoutImprovement

int maxIterationsWithoutImprovement

The maximum number of iterations to go without improvement before stopping

iterationsWithoutImprovement

int iterationsWithoutImprovement

The number of iterations since the last improvement.

population

Collection<E> population

The population of genomes.

initialPopulation

Collection<E> initialPopulation

The initial population of genomes

Class gov.sandia.cognition.learning.algorithm.genetic.ParallelizedGeneticAlgorithm extends GeneticAlgorithm<CostParametersType,GenomeType> implements Serializable

Serialized Fields

evaluateTasks

ArrayList<E> evaluateTasks

Parallel tasks that evaluate genome fitness

Package gov.sandia.cognition.learning.algorithm.genetic.reproducer

Class gov.sandia.cognition.learning.algorithm.genetic.reproducer.CrossoverReproducer extends Object implements Serializable

Serialized Fields

selector

Selector<GenomeType> selector

The selector to use to select the population.

crossoverFunction

CrossoverFunction<GenomeType> crossoverFunction

The crossover function to use.

Class gov.sandia.cognition.learning.algorithm.genetic.reproducer.MultiReproducer extends Object implements Serializable

Serialized Fields

reproducers

Collection<E> reproducers

The reproducers to use for reproducing.

Class gov.sandia.cognition.learning.algorithm.genetic.reproducer.MutationReproducer extends Object implements Serializable

Serialized Fields

perturber

Perturber<PerturbedType> perturber

The perturber to use for mutation.

selector

Selector<GenomeType> selector

The selector to use to select the population.

Class gov.sandia.cognition.learning.algorithm.genetic.reproducer.VectorizableCrossoverFunction extends AbstractRandomized implements Serializable

Serialized Fields

probabilityCrossover

double probabilityCrossover

Probability that an element in the child will come from vector2, and with probability (1-probabilityCrossover) the element will come from vector1. Thus, probabilityCrossover==0.0 means that all elements will come from vector1 and probabilityCrossover==1.0 means that all elements will come from vector2, probabilityCrossover==0.5 has maximum entropy in terms of where the elements of the child vector came from

Package gov.sandia.cognition.learning.algorithm.genetic.selector

Class gov.sandia.cognition.learning.algorithm.genetic.selector.AbstractSelector extends Object implements Serializable

Class gov.sandia.cognition.learning.algorithm.genetic.selector.TournamentSelector extends AbstractSelector<GenomeType> implements Serializable

Serialized Fields

percent

double percent

The percent of the population to select.

tournamentSize

int tournamentSize

The size of the tournament.

random

Random random

The random number generator to use.

Package gov.sandia.cognition.learning.algorithm.gradient

Class gov.sandia.cognition.learning.algorithm.gradient.GradientDescendableApproximator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

deltaSize

double deltaSize

Size of the finite-difference unit vectors, typically ~1e-5

function

VectorizableVectorFunction function

Internal VectorizableVectorFunction to consider

Package gov.sandia.cognition.learning.algorithm.hmm

Class gov.sandia.cognition.learning.algorithm.hmm.AbstractBaumWelchAlgorithm extends AbstractAnytimeBatchLearner<DataType,HiddenMarkovModel<ObservationType>> implements Serializable

Serialized Fields

distributionLearner

BatchLearner<DataType,ResultType> distributionLearner

Learner for the Distribution Functions of the HMM.

result

HiddenMarkovModel<ObservationType> result

Result of the Baum-Welch Algorithm

initialGuess

HiddenMarkovModel<ObservationType> initialGuess

Initial guess for the iterations.

lastLogLikelihood

double lastLogLikelihood

Last Log Likelihood of the iterations

reestimateInitialProbabilities

boolean reestimateInitialProbabilities

Flag to re-estimate the initial probability Vector.

Class gov.sandia.cognition.learning.algorithm.hmm.BaumWelchAlgorithm extends AbstractBaumWelchAlgorithm<ObservationType,Collection<? extends ObservationType>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.hmm.HiddenMarkovModel extends MarkovChain implements Serializable

Serialized Fields

emissionFunctions

Collection<E> emissionFunctions

The PDFs that emit symbols from each state.

Class gov.sandia.cognition.learning.algorithm.hmm.MarkovChain extends AbstractCloneableSerializable implements Serializable

Serialized Fields

initialProbability

Vector initialProbability

Initial probability Vector over the states. Each entry must be nonnegative and the Vector must sum to 1.

transitionProbability

Matrix transitionProbability

Transition probability matrix. The entry (i,j) is the probability of transition from state "j" to state "i". As a corollary, all entries in the Matrix must be nonnegative and the columns of the Matrix must sum to 1.

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelBaumWelchAlgorithm extends BaumWelchAlgorithm<ObservationType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelBaumWelchAlgorithm.DistributionEstimatorTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

weightedValues

ArrayList<E> weightedValues

Weighted values for the PDF estimator.

distributionLearner

BatchLearner<DataType,ResultType> distributionLearner

My copy of the PDF estimator.

gammas

ArrayList<E> gammas

Gammas used to weight the learner samples.

index

int index

Index into the gammas to pull the weights.

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel extends HiddenMarkovModel<ObservationType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel.ComputeTransitionsTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

alphan

Vector alphan

Alpha at time n.

betanp1

Vector betanp1

Alpha at time n.

bnp1

Vector bnp1

b at time n+1.

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel.LogLikelihoodTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

data

Collection<E> data

Data to compute the log-likelihood of

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel.NormalizeTransitionTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

A

Matrix A

Matrix to normalize.

j

int j

Column to normalize.

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel.ObservationLikelihoodTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

observations

Collection<E> observations

Observations

distributionFunction

ProbabilityFunction<DataType> distributionFunction

The PDF.

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel.StateObservationLikelihoodTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

alpha

Vector alpha

Alpha at time n.

beta

Vector beta

Beta at time n.

Class gov.sandia.cognition.learning.algorithm.hmm.ParallelHiddenMarkovModel.ViterbiTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

destinationState

int destinationState

Destination state for the Viterbi Recursion.

delta

Vector delta

Previous value of the Viterbi Recursion.

Package gov.sandia.cognition.learning.algorithm.minimization

Class gov.sandia.cognition.learning.algorithm.minimization.AbstractAnytimeFunctionMinimizer extends AbstractAnytimeBatchLearner<EvaluatorType extends Evaluator<? super InputType,? extends OutputType>,InputOutputPair<InputType,OutputType>> implements Serializable

Serialized Fields

tolerance

double tolerance

Tolerance of the minimization algorithm, must be >= 0.0

result

InputOutputPair<InputType,OutputType> result

Resulting minimum input-output pair

initialGuess

Object initialGuess

Initial guess of the minimization routine

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerBFGS extends FunctionMinimizerQuasiNewton implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerConjugateGradient extends AbstractAnytimeFunctionMinimizer<Vector,Double,DifferentiableEvaluator<? super Vector,Double,Vector>> implements Serializable

Serialized Fields

lineMinimizer

LineMinimizer<EvaluatorType extends Evaluator<Double,Double>> lineMinimizer

Work-horse algorithm that minimizes the function along a direction

lineFunction

DirectionalVectorToDifferentiableScalarFunction lineFunction

Function that maps a Evaluator onto a Evaluator using a set point, direction and scale factor

gradient

Vector gradient

Gradient at the current guess

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerDFP extends FunctionMinimizerQuasiNewton implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerDirectionSetPowell extends AbstractAnytimeFunctionMinimizer<Vector,Double,Evaluator<? super Vector,Double>> implements Serializable

Serialized Fields

lineMinimizer

LineMinimizer<EvaluatorType extends Evaluator<Double,Double>> lineMinimizer

Work-horse algorithm that minimizes the function along a direction

directionSet

List<E> directionSet

Matrix where the columns indicate the directions of minimization

lineFunction

DirectionalVectorToScalarFunction lineFunction

Function that maps a Evaluator onto a Evaluator using a set point, direction and scale factor

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerFletcherReeves extends FunctionMinimizerConjugateGradient implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerGradientDescent extends AbstractAnytimeFunctionMinimizer<Vector,Double,DifferentiableEvaluator<? super Vector,Double,Vector>> implements Serializable

Serialized Fields

learningRate

double learningRate

The learning rate (or step size), must be (0,1], typically ~0.1

momentum

double momentum

The momentum rate, must be [0,1), typically ~0.8

previousDelta

Vector previousDelta

Previous input change, used for adding momentum

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerLiuStorey extends FunctionMinimizerConjugateGradient implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerNelderMead extends AbstractAnytimeFunctionMinimizer<Vector,Double,Evaluator<? super Vector,Double>> implements Serializable

Serialized Fields

simplex

ArrayList<E> simplex

InputOutputPairs that define the simplex

simplexInputSum

Vector simplexInputSum

A running sum of the inputs of the simplex, used for reflection

lineFunction

DirectionalVectorToScalarFunction lineFunction

Function that defines the simplex growth direction

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerPolakRibiere extends FunctionMinimizerConjugateGradient implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.FunctionMinimizerQuasiNewton extends AbstractAnytimeFunctionMinimizer<Vector,Double,DifferentiableEvaluator<? super Vector,Double,Vector>> implements Serializable

Serialized Fields

lineMinimizer

LineMinimizer<EvaluatorType extends Evaluator<Double,Double>> lineMinimizer

Work-horse algorithm that minimizes the function along a direction

lineFunction

DirectionalVectorToDifferentiableScalarFunction lineFunction

Function that maps a Evaluator onto a Evaluator using a set point, direction and scale factor

hessianInverse

Matrix hessianInverse

Estimated inverse of the Hessian (second derivative)

gradient

Vector gradient

Gradient at the current guess

dimensionality

int dimensionality

The dimensionality of the input space

Package gov.sandia.cognition.learning.algorithm.minimization.line

Class gov.sandia.cognition.learning.algorithm.minimization.line.AbstractAnytimeLineMinimizer extends AbstractAnytimeFunctionMinimizer<Double,Double,EvaluatorType extends Evaluator<Double,Double>> implements Serializable

Serialized Fields

bracket

LineBracket bracket

LineBracket bounding a local minimum.

validBracket

boolean validBracket

Flag indicating if the algorithm has already found a valid bracket on a local minimum.

interpolator

LineBracketInterpolator<EvaluatorType extends Evaluator<Double,Double>> interpolator

Type of algorithm to fit data points and find an interpolated minimum to the known points.

initialGuessFunctionValue

Double initialGuessFunctionValue

Function value at the initialGuess, may be null.

initialGuessSlope

Double initialGuessSlope

Function slope at the initialGuess, may be null.

Class gov.sandia.cognition.learning.algorithm.minimization.line.DirectionalVectorToDifferentiableScalarFunction extends DirectionalVectorToScalarFunction implements Serializable

Serialized Fields

lastGradient

InputOutputPair<InputType,OutputType> lastGradient

Last gradient information

Class gov.sandia.cognition.learning.algorithm.minimization.line.DirectionalVectorToScalarFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Serialized Fields

vectorOffset

Vector vectorOffset

Vector offset to scale in the specified direction

direction

Vector direction

Directional vector for optimization

vectorScalarFunction

Evaluator<InputType,OutputType> vectorScalarFunction

Function that maps a Vector onto a Double

lastEvaluation

DefaultInputOutputPair<InputType,OutputType> lastEvaluation

Cache for the last input/output that was evaluated, so that we can avoid using it again in the differentiate method.

Class gov.sandia.cognition.learning.algorithm.minimization.line.InputOutputSlopeTriplet extends DefaultInputOutputPair<Double,Double> implements Serializable

Serialized Fields

slope

Double slope

Slope (first derivative) at the given point

Class gov.sandia.cognition.learning.algorithm.minimization.line.LineBracket extends AbstractCloneableSerializable implements Serializable

Serialized Fields

lowerBound

InputOutputSlopeTriplet lowerBound

Lower bound of the bracket.

upperBound

InputOutputSlopeTriplet upperBound

Upper bound of the bracket.

otherPoint

InputOutputSlopeTriplet otherPoint

Another (optional) point associated with the bracket.

Class gov.sandia.cognition.learning.algorithm.minimization.line.LineMinimizerBacktracking extends AbstractAnytimeLineMinimizer<Evaluator<Double,Double>> implements Serializable

Serialized Fields

sufficientDecrease

double sufficientDecrease

Sufficient decrease condition, must be (0,1). Smaller values (0.1) result in more accurate searches, larger values (0.9) tend to be easier to satisfy.

geometricDecrease

double geometricDecrease

Amount to decrease the step amount each iteration.

numericalDerivative

boolean numericalDerivative

Flag whether or not to use the numerical differentiation.

stepValue

double stepValue

Current value of the step size.

Class gov.sandia.cognition.learning.algorithm.minimization.line.LineMinimizerDerivativeBased extends AbstractAnytimeLineMinimizer<DifferentiableUnivariateScalarFunction> implements Serializable

Serialized Fields

minFunctionValue

double minFunctionValue

Minimum value of the target function. In other words, the user will accept a solution less than or equal to minFunctionValue. For many applications 0.0 is a likely candidate (for cost functions, metrics, least squares, etc.)

direction

double direction

Direction of the search. Because Fletcher assumes the slope of the initialGuess is less than 0.0, we have to flip around the direction of search if the initial guess has positive slope. Thus, direction=1.0 means that the initial slope was negative, while direction=-1.0 means that the initial slope was positive.

internalFunction

LineMinimizerDerivativeBased.InternalFunction internalFunction

Internal function used to map/remap/unmap the search direction.

wolfe

WolfeConditions wolfe

The Wolfe conditions define approximate line search stopping criteria.

maxX

double maxX

Maximum value of x in the search space. That is, the minimizer will not be greater than maxX.

Class gov.sandia.cognition.learning.algorithm.minimization.line.LineMinimizerDerivativeBased.InternalFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.line.LineMinimizerDerivativeFree extends AbstractAnytimeLineMinimizer<Evaluator<Double,Double>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.line.WolfeConditions extends AbstractCloneableSerializable implements Serializable

Serialized Fields

originalPoint

InputOutputSlopeTriplet originalPoint

Original point to store, slope must be less than 0.0.

slopeCondition

double slopeCondition

Slope condition parameter for the Goldstein condition, must be less than curvatureCondition and on the interval (0,1).

curvatureCondition

double curvatureCondition

Curvature condition for the curvature condition, must be greater than slopeCondition and on the interval (0,1).

Package gov.sandia.cognition.learning.algorithm.minimization.line.interpolator

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.AbstractLineBracketInterpolator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

tolerance

double tolerance

Tolerance of the interpolator to collinear or identical points

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.AbstractLineBracketInterpolatorPolynomial extends AbstractLineBracketInterpolator<EvaluatorType extends Evaluator<Double,Double>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.LineBracketInterpolatorBrent extends AbstractLineBracketInterpolator<Evaluator<Double,Double>> implements Serializable

Serialized Fields

parabolicInterpolator

LineBracketInterpolatorParabola parabolicInterpolator

Non-slope based parabolic interpolator.

goldenInterpolator

LineBracketInterpolatorGoldenSection goldenInterpolator

Golden-section step interpolator.

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.LineBracketInterpolatorGoldenSection extends AbstractLineBracketInterpolator<Evaluator<Double,Double>> implements Serializable

Serialized Fields

linearInterpolator

LineBracketInterpolatorLinear linearInterpolator

Back-up interpolator using the secant method.

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.LineBracketInterpolatorHermiteCubic extends AbstractLineBracketInterpolatorPolynomial<DifferentiableUnivariateScalarFunction> implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.LineBracketInterpolatorHermiteParabola extends AbstractLineBracketInterpolatorPolynomial<DifferentiableUnivariateScalarFunction> implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.LineBracketInterpolatorLinear extends AbstractLineBracketInterpolatorPolynomial<Evaluator<Double,Double>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.minimization.line.interpolator.LineBracketInterpolatorParabola extends AbstractLineBracketInterpolatorPolynomial<Evaluator<Double,Double>> implements Serializable

Package gov.sandia.cognition.learning.algorithm.nearest

Class gov.sandia.cognition.learning.algorithm.nearest.AbstractKNearestNeighbor extends AbstractNearestNeighbor<InputType,OutputType> implements Serializable

Serialized Fields

k

int k

Number of neighbors to consider, must be greater than zero

averager

Summarizer<DataType,SummaryType> averager

Creates a single object from a collection of data

Class gov.sandia.cognition.learning.algorithm.nearest.AbstractNearestNeighbor extends DefaultDivergenceFunctionContainer<InputType,InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.nearest.KNearestNeighborExhaustive extends AbstractKNearestNeighbor<InputType,OutputType> implements Serializable

Serialized Fields

data

Collection<E> data

Underlying data for the classifier

Class gov.sandia.cognition.learning.algorithm.nearest.KNearestNeighborExhaustive.Learner extends KNearestNeighborExhaustive<InputType,OutputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.nearest.KNearestNeighborExhaustive.Neighbor extends AbstractCloneableSerializable implements Serializable

Serialized Fields

value

Object value

Output value.

divergence

double divergence

Divergence associated with this value.

Class gov.sandia.cognition.learning.algorithm.nearest.KNearestNeighborKDTree extends AbstractKNearestNeighbor<InputType extends Vectorizable,OutputType> implements Serializable

Serialized Fields

data

KDTree<VectorType extends Vectorizable,DataType,PairType extends Pair<? extends VectorType,DataType>> data

KDTree that holds the data to search for neighbors.

Class gov.sandia.cognition.learning.algorithm.nearest.KNearestNeighborKDTree.Learner extends KNearestNeighborKDTree<InputType extends Vectorizable,OutputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.nearest.NearestNeighborExhaustive extends AbstractNearestNeighbor<InputType,OutputType> implements Serializable

Serialized Fields

data

LinkedList<E> data

The data that nearest-neighbor is performed over.

Class gov.sandia.cognition.learning.algorithm.nearest.NearestNeighborExhaustive.Learner extends DefaultDivergenceFunctionContainer<InputType,InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.nearest.NearestNeighborKDTree extends AbstractNearestNeighbor<InputType extends Vectorizable,OutputType> implements Serializable

Serialized Fields

data

KDTree<VectorType extends Vectorizable,DataType,PairType extends Pair<? extends VectorType,DataType>> data

KDTree that holds the data to search for neighbors.

Class gov.sandia.cognition.learning.algorithm.nearest.NearestNeighborKDTree.Learner extends NearestNeighborKDTree<InputType extends Vectorizable,OutputType> implements Serializable

Package gov.sandia.cognition.learning.algorithm.pca

Class gov.sandia.cognition.learning.algorithm.pca.AbstractPrincipalComponentsAnalysis extends AbstractCloneableSerializable implements Serializable

Serialized Fields

numComponents

int numComponents

Number of components to extract from the data, must be greater than zero

result

PrincipalComponentsAnalysisFunction result

Vector function that maps the input space onto a numComponents-dimension Vector representing the directions of maximal variance (information gain). The i-th row in the matrix approximates the i-th column of the "U" matrix of the Singular Value Decomposition.

Class gov.sandia.cognition.learning.algorithm.pca.GeneralizedHebbianAlgorithm extends AbstractAnytimeBatchLearner<Collection<Vector>,PrincipalComponentsAnalysisFunction> implements Serializable

Serialized Fields

learningRate

double learningRate

Learning rate, or step size, (0,1], typically ~0.1

numComponents

int numComponents

Number of components to extract from the data, must be greater than zero

result

PrincipalComponentsAnalysisFunction result

Vector function that maps the input space onto a numComponents-dimension Vector representing the directions of maximal variance (information gain). The i-th row in the matrix approximates the i-th column of the "U" matrix of the Singular Value Decomposition.

components

ArrayList<E> components

Components that have been extracted from the input data, each component has the same dimensions as the input data and the size of the ArrayList is numComponents

mean

Vector mean

Sample mean of the training data. This is subtracted from the training data before GHA is executed.

minChange

double minChange

Minimum change below which to stop iterating, greater than or equal to zero, typically 1e-10

Class gov.sandia.cognition.learning.algorithm.pca.KernelPrincipalComponentsAnalysis extends DefaultKernelContainer<DataType> implements Serializable

Serialized Fields

componentCount

int componentCount

The number of components to create from the analysis. Must be positive.

centerData

boolean centerData

Whether or not the data should be centered before doing KPCA.

Class gov.sandia.cognition.learning.algorithm.pca.KernelPrincipalComponentsAnalysis.Function extends DefaultKernelContainer<DataType> implements Serializable

Serialized Fields

data

List<E> data

The data that the KPCA was performed over. Each one corresponds to a column in the components matrix.

components

Matrix components

The matrix of components for the function. The number of rows is the dimensionality of the reduction. The number of columns is equal to the number of data points that the KPCA was done over.

centerData

boolean centerData

A flag indicating if the incoming data needs to be centered or not. Unless the data is being pre-centered, this should be true.

kernelMatrix

Matrix kernelMatrix

The kernel matrix for all the data the KPCA was done over. It is a square matrix whose size is equal to the data.

Class gov.sandia.cognition.learning.algorithm.pca.PrincipalComponentsAnalysisFunction extends AbstractCloneableSerializable implements Serializable

Serialized Fields

mean

Vector mean

Sample mean of the data, which is subtracted from input data to yield zero-mean inputs

dimensionReducer

MultivariateDiscriminant dimensionReducer

Function that maps a high-dimension input space onto a (hopefully) simple low-dimensional output space capturing the directions of maximum variance (information gain)

Class gov.sandia.cognition.learning.algorithm.pca.ThinSingularValueDecomposition extends AbstractPrincipalComponentsAnalysis implements Serializable

Package gov.sandia.cognition.learning.algorithm.perceptron

Class gov.sandia.cognition.learning.algorithm.perceptron.AbstractKernelizableBinaryCategorizerOnlineLearner extends AbstractOnlineLinearBinaryCategorizerLearner implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.AbstractLinearCombinationOnlineLearner extends AbstractKernelizableBinaryCategorizerOnlineLearner implements Serializable

Serialized Fields

updateBias

boolean updateBias

An option controlling whether or not the bias is updated or not. Individual algorithm implementations choose the default value for this.

Class gov.sandia.cognition.learning.algorithm.perceptron.AbstractOnlineLinearBinaryCategorizerLearner extends AbstractSupervisedBatchAndIncrementalLearner<Vectorizable,Boolean,LinearBinaryCategorizer> implements Serializable

Serialized Fields

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The factory to create weight vectors.

Class gov.sandia.cognition.learning.algorithm.perceptron.AggressiveRelaxedOnlineMaximumMarginAlgorithm extends AbstractKernelizableBinaryCategorizerOnlineLearner implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.Ballseptron extends AbstractKernelizableBinaryCategorizerOnlineLearner implements Serializable

Serialized Fields

radius

double radius

The radius enforced by the algorithm.

Class gov.sandia.cognition.learning.algorithm.perceptron.BatchMultiPerceptron extends AbstractAnytimeSupervisedBatchLearner<Vectorizable,CategoryType,LinearMultiCategorizer<CategoryType>> implements Serializable

Serialized Fields

minMargin

double minMargin

The minimum margin to enforce. Must be non-negative.

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The factory to create weight vectors.

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineBinaryMarginInfusedRelaxedAlgorithm extends AbstractLinearCombinationOnlineLearner implements Serializable

Serialized Fields

minMargin

double minMargin

The minimum margin to enforce. Must be non-negative.

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineMultiPerceptron extends AbstractBatchAndIncrementalLearner<InputOutputPair<? extends Vectorizable,CategoryType>,LinearMultiCategorizer<CategoryType>> implements Serializable

Serialized Fields

minMargin

double minMargin

The minimum margin to enforce. Must be non-negative.

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The factory to create weight vectors.

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineMultiPerceptron.ProportionalUpdate extends OnlineMultiPerceptron<CategoryType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineMultiPerceptron.UniformUpdate extends OnlineMultiPerceptron<CategoryType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlinePassiveAggressivePerceptron extends AbstractLinearCombinationOnlineLearner implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlinePassiveAggressivePerceptron.AbstractSoftMargin extends OnlinePassiveAggressivePerceptron implements Serializable

Serialized Fields

aggressiveness

double aggressiveness

The aggressiveness parameter (C), which is the trade-off between aggressive updating to meet an incorrect example and keeping history around.

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlinePassiveAggressivePerceptron.LinearSoftMargin extends OnlinePassiveAggressivePerceptron.AbstractSoftMargin implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlinePassiveAggressivePerceptron.QuadraticSoftMargin extends OnlinePassiveAggressivePerceptron.AbstractSoftMargin implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlinePerceptron extends AbstractLinearCombinationOnlineLearner implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineRampPassiveAggressivePerceptron extends OnlinePassiveAggressivePerceptron.QuadraticSoftMargin implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineShiftingPerceptron extends AbstractLinearCombinationOnlineLearner implements Serializable

Serialized Fields

lambda

double lambda

The lambda parameter for controlling how much shifting occurs.

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineShiftingPerceptron.LinearResult extends LinearBinaryCategorizer implements Serializable

Serialized Fields

errorCount

long errorCount

The number of errors made by the categorizer so far.

Class gov.sandia.cognition.learning.algorithm.perceptron.OnlineVotedPerceptron extends AbstractSupervisedBatchAndIncrementalLearner<Vectorizable,Boolean,WeightedBinaryEnsemble<Vectorizable,LinearBinaryCategorizer>> implements Serializable

Serialized Fields

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The factory to create weight vectors.

Class gov.sandia.cognition.learning.algorithm.perceptron.Perceptron extends AbstractAnytimeSupervisedBatchLearner<Vectorizable,Boolean,LinearBinaryCategorizer> implements Serializable

Serialized Fields

marginPositive

double marginPositive

The positive margin to enforce.

marginNegative

double marginNegative

The negative margin to enforce.

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The VectorFactory to use to create vectors.

result

LinearBinaryCategorizer result

The result categorizer.

errorCount

int errorCount

The number of errors on the most recent iteration.

Class gov.sandia.cognition.learning.algorithm.perceptron.RelaxedOnlineMaximumMarginAlgorithm extends AbstractKernelizableBinaryCategorizerOnlineLearner implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.Winnow extends AbstractOnlineLinearBinaryCategorizerLearner implements Serializable

Serialized Fields

weightUpdate

double weightUpdate

The amount of the weight update (alpha). Must be greater than 1.

demoteToZero

boolean demoteToZero

An option to demote to zero.

weightUpdateInverse

double weightUpdateInverse

The cached value of the inverse of weight update (commonly alpha or 1 + epsilon).

Package gov.sandia.cognition.learning.algorithm.perceptron.kernel

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.AbstractOnlineBudgetedKernelBinaryCategorizerLearner extends AbstractOnlineKernelBinaryCategorizerLearner<InputType> implements Serializable

Serialized Fields

budget

int budget

The budget of the number of examples to keep. Must be positive.

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.AbstractOnlineKernelBinaryCategorizerLearner extends AbstractSupervisedBatchAndIncrementalLearner<InputType,Boolean,DefaultKernelBinaryCategorizer<InputType>> implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The kernel to use.

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Forgetron extends AbstractOnlineBudgetedKernelBinaryCategorizerLearner<InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Forgetron.Basic extends AbstractOnlineBudgetedKernelBinaryCategorizerLearner<InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Forgetron.Greedy extends Forgetron<InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Forgetron.Result extends DefaultKernelBinaryCategorizer<InputType> implements Serializable

Serialized Fields

errorCount

long errorCount

The number of errors that the categorizer has made in the learning step.

q

double q

The value of Q for the algorithm.

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.KernelAdatron extends AbstractAnytimeSupervisedBatchLearner<InputType,Boolean,KernelBinaryCategorizer<InputType,DefaultWeightedValue<InputType>>> implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The kernel to use.

result

KernelBinaryCategorizer<InputType,EntryType extends WeightedValue<? extends InputType>> result

The result categorizer.

errorCount

int errorCount

The number of errors on the most recent iteration.

supportsMap

LinkedHashMap<K,V> supportsMap

The mapping of weight objects to non-zero weighted examples (support vectors).

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.KernelBinaryCategorizerOnlineLearnerAdapter extends AbstractOnlineKernelBinaryCategorizerLearner<InputType> implements Serializable

Serialized Fields

learner

KernelizableBinaryCategorizerOnlineLearner learner

The wrapped kernelizable learner.

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.KernelPerceptron extends AbstractAnytimeSupervisedBatchLearner<InputType,Boolean,DefaultKernelBinaryCategorizer<InputType>> implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The kernel to use.

marginPositive

double marginPositive

The positive margin to enforce.

marginNegative

double marginNegative

The negative margin to enforce.

result

DefaultKernelBinaryCategorizer<InputType> result

The result categorizer.

errorCount

int errorCount

The number of errors on the most recent iteration.

supportsMap

LinkedHashMap<K,V> supportsMap

The mapping of weight objects to non-zero weighted examples (support vectors).

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.OnlineKernelPerceptron extends AbstractOnlineKernelBinaryCategorizerLearner<InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.OnlineKernelRandomizedBudgetPerceptron extends AbstractOnlineBudgetedKernelBinaryCategorizerLearner<InputType> implements Serializable

Serialized Fields

random

Random random

The random number generator.

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Projectron extends AbstractOnlineKernelBinaryCategorizerLearner<InputType> implements Serializable

Serialized Fields

eta

double eta

The eta parameter, which ends up controlling the number of supports created. Must be non-negative.

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Projectron.LinearSoftMargin extends Projectron<InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.RemoveOldestKernelPerceptron extends AbstractOnlineBudgetedKernelBinaryCategorizerLearner<InputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.perceptron.kernel.Stoptron extends AbstractOnlineBudgetedKernelBinaryCategorizerLearner<InputType> implements Serializable

Package gov.sandia.cognition.learning.algorithm.regression

Class gov.sandia.cognition.learning.algorithm.regression.AbstractMinimizerBasedParameterCostMinimizer extends AnytimeAlgorithmWrapper<ResultType extends VectorizableVectorFunction,FunctionMinimizer<Vector,Double,? super EvaluatorType extends Evaluator<? super Vector,? extends Double>>> implements Serializable

Serialized Fields

objectToOptimize

VectorizableVectorFunction objectToOptimize

Object that is being optimized.

result

VectorizableVectorFunction result

Resulting value

costFunction

SupervisedCostFunction<InputType,TargetType> costFunction

Cost function to compute the cost of objectToOptimize

Class gov.sandia.cognition.learning.algorithm.regression.AbstractParameterCostMinimizer extends AbstractAnytimeSupervisedBatchLearner<Vector,Vector,ResultType extends VectorizableVectorFunction> implements Serializable

Serialized Fields

objectToOptimize

VectorizableVectorFunction objectToOptimize

GradientDescendable whose parameters result minimize the cost function

result

VectorizableVectorFunction result

Result to return

tolerance

double tolerance

Stopping criterion for the algorithm, typically ~1e-5

costFunction

SupervisedCostFunction<InputType,TargetType> costFunction

Cost function that computes the cost of the object to optimize

resultCost

Double resultCost

Cost of the result

Class gov.sandia.cognition.learning.algorithm.regression.FletcherXuHybridEstimation extends LeastSquaresEstimator implements Serializable

Serialized Fields

lineMinimizer

LineMinimizer<EvaluatorType extends Evaluator<Double,Double>> lineMinimizer

Workhorse algorithm that finds the minimum along a particular direction

reductionTest

double reductionTest

Reduction test for switching between BFGS and Levenberg-Marquardt, must be [0,1]. Lower values result in more Levenberg-Marquardt steps, larger values result in more BFGS steps.

dampingFactorDivisor

double dampingFactorDivisor

Amount to modify the damping factor, typically 2.0 or 10.0

lastCost

SumSquaredErrorCostFunction.Cache lastCost

Last value of the parameter cost

lineFunction

DirectionalVectorToDifferentiableScalarFunction lineFunction

Function that maps a Evaluator onto a Evaluator using a set point, direction and scale factor

hessianInverse

Matrix hessianInverse

Estimated inverse of the Hessian (second derivative)

dampingFactor

double dampingFactor

Damping factor for the Levenberg-Marquardt ridge regression

Class gov.sandia.cognition.learning.algorithm.regression.GaussNewtonAlgorithm extends LeastSquaresEstimator implements Serializable

Serialized Fields

lineMinimizer

LineMinimizer<EvaluatorType extends Evaluator<Double,Double>> lineMinimizer

Workhorse algorithm that finds the minimum along a particular direction

lineFunction

DirectionalVectorToDifferentiableScalarFunction lineFunction

Function that maps a Evaluator onto a Evaluator using a set point, direction and scale factor

Class gov.sandia.cognition.learning.algorithm.regression.KernelBasedIterativeRegression extends AbstractAnytimeSupervisedBatchLearner<InputType,Double,KernelScalarFunction<InputType>> implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The kernel to use.

minSensitivity

double minSensitivity

The bound on sensitivity.

result

KernelScalarFunction<InputType> result

The result categorizer.

errorCount

int errorCount

The number of errors on the most recent iteration.

Class gov.sandia.cognition.learning.algorithm.regression.KernelWeightedRobustRegression extends AbstractAnytimeSupervisedBatchLearner<InputType,OutputType,Evaluator<? super InputType,? extends OutputType>> implements Serializable

Serialized Fields

result

Evaluator<InputType,OutputType> result

DecoupledVectorFunction that is being optimized

iterationLearner

SupervisedBatchLearner<InputType,OutputType,ResultType extends Evaluator<? super InputType,? extends OutputType>> iterationLearner

Internal learning algorithm that computes optimal solutions given the current weightedData. The iterationLearner should operate on WeightedInputOutputPairs (we have a hard time enforcing this, as many learning algorithms operate both on InputOutputPairs and WeightedInputOutputPairs)

kernelWeightingFunction

Kernel<InputType> kernelWeightingFunction

Kernel function that provides the weighting for the estimate error, generally the Kernel should weight accurate estimates higher than inaccurate estimates.

tolerance

double tolerance

Tolerance before stopping the algorithm

weightedData

ArrayList<E> weightedData

Weighted copy of the data

Class gov.sandia.cognition.learning.algorithm.regression.LeastSquaresEstimator extends AbstractParameterCostMinimizer<GradientDescendable,SumSquaredErrorCostFunction> implements Serializable

Class gov.sandia.cognition.learning.algorithm.regression.LevenbergMarquardtEstimation extends LeastSquaresEstimator implements Serializable

Serialized Fields

iterationsWithoutImprovement

int iterationsWithoutImprovement

Number of sequential unsuccessful iterations without a cost-reducing step

maxIterationsWithoutImprovement

int maxIterationsWithoutImprovement

Maximum number of iterations without improvement before stopping

dampingFactor

double dampingFactor

Current damping factor for the ridge regression

dampingFactorDivisor

double dampingFactorDivisor

Divisor of the damping factor on a successful iteration, must be greater then 1.0, typically ~10.0

bestParameters

Vector bestParameters

Parameters used to generate the lastCost

bestParametersCost

SumSquaredErrorCostFunction.Cache bestParametersCost

Cost associated with lastParameters

Class gov.sandia.cognition.learning.algorithm.regression.LinearBasisRegression extends AbstractCloneableSerializable implements Serializable

Serialized Fields

inputToVectorMap

Evaluator<InputType,OutputType> inputToVectorMap

Function that maps the InputType to a Vector

usePseudoInverse

boolean usePseudoInverse

Flag to use a pseudoinverse. True to use the expensive, but more accurate, pseudoinverse routine. False uses a very fast, but numerically less stable LU solver. Default value is "true".

Class gov.sandia.cognition.learning.algorithm.regression.LinearRegression extends AbstractCloneableSerializable implements Serializable

Serialized Fields

usePseudoInverse

boolean usePseudoInverse

Flag to use a pseudoinverse. True to use the expensive, but more accurate, pseudoinverse routine. False uses a very fast, but numerically less stable LU solver. Default value is "true".

regularization

double regularization

L2 ridge regularization term, must be nonnegative, a value of zero is equivalent to unregularized regression.

Class gov.sandia.cognition.learning.algorithm.regression.LinearRegression.Statistic extends AbstractConfidenceStatistic implements Serializable

Serialized Fields

chiSquare

double chiSquare

Gets the value of the chi-square variable, Total weighted sum-squared error between the targets and estimates

rootMeanSquaredError

double rootMeanSquaredError

Root mean-squared error of the targets and estimates

meanL1Error

double meanL1Error

Average L1-norm error (absolute value difference) between the targets and estimates

targetEstimateCorrelation

double targetEstimateCorrelation

Pearson Correlation between the targets and estimates, [-1,1]

unpredictedErrorFraction

double unpredictedErrorFraction

Fraction of variance unaccounted for in the predictions, [0,1]

numSamples

int numSamples

Number of samples used to create the Regression

numParameters

int numParameters

Number of parameters in the learned approximator

degreesOfFreedom

double degreesOfFreedom

Degrees of freedom in the Regression = numSamples-numParameters

Class gov.sandia.cognition.learning.algorithm.regression.LocallyWeightedFunction.Learner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

Kernel that provides the weights between an input and each sample in the input dataset

learner

SupervisedBatchLearner<InputType,OutputType,ResultType extends Evaluator<? super InputType,? extends OutputType>> learner

Learner that takes the Collection of WeightedInputOutputPairs from the Kernel reweighting and creates a local function approximation at the given input. I would strongly recommend using fast or closed-form learners for this.

Class gov.sandia.cognition.learning.algorithm.regression.LogisticRegression extends AbstractAnytimeSupervisedBatchLearner<Vectorizable,Double,LogisticRegression.Function> implements Serializable

Serialized Fields

objectToOptimize

LogisticRegression.Function objectToOptimize

The object to optimize, used as a factory on successive runs of the algorithm.

result

LogisticRegression.Function result

Return value from the algorithm

tolerance

double tolerance

Tolerance change in weights before stopping

regularization

double regularization

L2 ridge regularization term, must be nonnegative, a value of zero is equivalent to unregularized regression.

Class gov.sandia.cognition.learning.algorithm.regression.LogisticRegression.Function extends CompositeEvaluatorPair<Vectorizable,Double,Double> implements Serializable

Class gov.sandia.cognition.learning.algorithm.regression.MultivariateLinearRegression extends AbstractCloneableSerializable implements Serializable

Serialized Fields

usePseudoInverse

boolean usePseudoInverse

Flag to use a pseudoinverse. True to use the expensive, but more accurate, pseudoinverse routine. False uses a very fast, but numerically less stable LU solver. Default value is "true".

regularization

double regularization

L2 ridge regularization term, must be nonnegative, a value of zero is equivalent to unregularized regression.

Class gov.sandia.cognition.learning.algorithm.regression.ParameterDerivativeFreeCostMinimizer extends AbstractMinimizerBasedParameterCostMinimizer<VectorizableVectorFunction,DifferentiableEvaluator<Vector,Double,Vector>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.regression.ParameterDerivativeFreeCostMinimizer.ParameterCostEvaluatorDerivativeFree extends AbstractCloneableSerializable implements Serializable

Serialized Fields

internalFunction

VectorizableVectorFunction internalFunction

Object that we're tweaking the parameters of.

costFunction

SupervisedCostFunction<InputType,TargetType> costFunction

Cost function against which to evaluate the cost of the object.

Class gov.sandia.cognition.learning.algorithm.regression.ParameterDifferentiableCostMinimizer extends AbstractMinimizerBasedParameterCostMinimizer<GradientDescendable,DifferentiableEvaluator<Vector,Double,Vector>> implements Serializable

Class gov.sandia.cognition.learning.algorithm.regression.ParameterDifferentiableCostMinimizer.ParameterCostEvaluatorDerivativeBased extends AbstractCloneableSerializable implements Serializable

Serialized Fields

internalFunction

GradientDescendable internalFunction

Object that we're tweaking the parameters of.

costFunction

DifferentiableCostFunction costFunction

Cost function against which to evaluate the cost of the object.

Class gov.sandia.cognition.learning.algorithm.regression.UnivariateLinearRegression extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.algorithm.root

Class gov.sandia.cognition.learning.algorithm.root.AbstractBracketedRootFinder extends AbstractRootFinder implements Serializable

Serialized Fields

bracketer

RootBracketer bracketer

Root-bracketing algorithm.

rootBracket

LineBracket rootBracket

Gets the bracket on the root location. That is, a range of input values where a root is guaranteed to lay between.

Class gov.sandia.cognition.learning.algorithm.root.AbstractRootFinder extends AbstractAnytimeBatchLearner<Evaluator<Double,Double>,InputOutputPair<Double,Double>> implements Serializable

Serialized Fields

tolerance

double tolerance

Tolerance, where tolerances closer to zero are more accurate, and larger tolerances are less accurate. In any case, tolerance must be greater than or equal to zero.

initialGuess

double initialGuess

Initial guess of the root location.

Class gov.sandia.cognition.learning.algorithm.root.MinimizerBasedRootFinder extends AnytimeAlgorithmWrapper<InputOutputPair<Double,Double>,LineMinimizer<Evaluator<Double,Double>>> implements Serializable

Serialized Fields

internalFunction

gov.sandia.cognition.learning.algorithm.root.MinimizerBasedRootFinder.MinimizationFunction internalFunction

Internal function.

Class gov.sandia.cognition.learning.algorithm.root.RootBracketExpander extends AbstractAnytimeBatchLearner<Evaluator<Double,Double>,LineBracket> implements Serializable

Serialized Fields

bracket

LineBracket bracket

Bracket on the root location.

initialGuess

double initialGuess

Initial guess of the root's location.

Class gov.sandia.cognition.learning.algorithm.root.RootFinderBisectionMethod extends AbstractBracketedRootFinder implements Serializable

Class gov.sandia.cognition.learning.algorithm.root.RootFinderFalsePositionMethod extends AbstractBracketedRootFinder implements Serializable

Class gov.sandia.cognition.learning.algorithm.root.RootFinderNewtonsMethod extends AbstractRootFinder implements Serializable

Serialized Fields

result

DefaultInputOutputPair<InputType,OutputType> result

Resulting estimated location of the root.

dfdx

DifferentiableEvaluator<InputType,OutputType,DerivativeType> dfdx

Internal Function variable from which we will pull the derivative. If the Evaluator is a DifferentiableEvaluator, then we will use that. Otherwise, we set up an approximation to the derivative.

stepMultiplier

double stepMultiplier

Multiplier of the current step.

Class gov.sandia.cognition.learning.algorithm.root.RootFinderRiddersMethod extends AbstractBracketedRootFinder implements Serializable

Class gov.sandia.cognition.learning.algorithm.root.RootFinderSecantMethod extends AbstractBracketedRootFinder implements Serializable

Serialized Fields

previousPoint

InputOutputSlopeTriplet previousPoint

Class gov.sandia.cognition.learning.algorithm.root.SolverFunction extends AbstractUnivariateScalarFunction implements Serializable

Serialized Fields

target

double target

internalFunction value to search for.

internalFunction

Evaluator<InputType,OutputType> internalFunction

The internal function to use.

Package gov.sandia.cognition.learning.algorithm.svm

Class gov.sandia.cognition.learning.algorithm.svm.PrimalEstimatedSubGradient extends AbstractAnytimeSupervisedBatchLearner<Vectorizable,Boolean,LinearBinaryCategorizer> implements Serializable

Serialized Fields

sampleSize

int sampleSize

The sample size requested by the user. The actual sample size may be less than this in the case that the sample size is larger than the amount of data given in the training set.

regularizationWeight

double regularizationWeight

The weight assigned to the regularization term in the algorithm, which is often represented as lambda.

random

Random random

The random number generator to use.

Class gov.sandia.cognition.learning.algorithm.svm.SequentialMinimalOptimization extends AbstractAnytimeSupervisedBatchLearner<InputType,Boolean,KernelBinaryCategorizer<InputType,DefaultWeightedValue<InputType>>> implements Serializable

Serialized Fields

maxPenalty

double maxPenalty

The maximum penalty parameter (C). Must be greater than 0.0.

errorTolerance

double errorTolerance

The error tolerance for the algorithm. Must be non-negative. Also known as "tol" or "tolerance".

effectiveZero

double effectiveZero

The effective value for zero to use in the computation to deal with numerical imprecision. Must be a non-negative number. Typically a very small value. Also sometimes known as epsilon.

kernelCacheSize

int kernelCacheSize

The size of the kernel cache, which is the number of kernel computations to keep cached. May be 0 to indicate that the kernel cache should not be used.

random

Random random

The random number generator to use.

kernel

Kernel<InputType> kernel

The kernel to use.

Class gov.sandia.cognition.learning.algorithm.svm.SuccessiveOverrelaxation extends AbstractAnytimeSupervisedBatchLearner<InputType,Boolean,KernelBinaryCategorizer<InputType,DefaultWeightedValue<InputType>>> implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The kernel to use.

maxWeight

double maxWeight

The maximum weight for a support vector. Must be greater than zero.

overrelaxation

double overrelaxation

The overrelaxation parameter. Must be in (0, 2), exclusive.

minChange

double minChange

The minimum change to allow for the algorithm to keep going. If the Total change is below this, then the algorithm will stop. Must be greater than zero.

result

KernelBinaryCategorizer<InputType,EntryType extends WeightedValue<? extends InputType>> result

The result categorizer.

totalChange

double totalChange

The total change on the most recent pass.

entries

ArrayList<E> entries

The entry information that the algorithm keeps.

supportsMap

LinkedHashMap<K,V> supportsMap

The mapping of weight objects to non-zero weighted examples (support vectors).

Class gov.sandia.cognition.learning.algorithm.svm.SuccessiveOverrelaxation.Entry extends DefaultWeightedValue<InputType> implements Serializable

Serialized Fields

example

InputOutputPair<InputType,OutputType> example

The example the data pertains to.

output

boolean output

The output represented as a raw boolean, to enforce that the label exists.

outputDouble

double outputDouble

The output converted to a double form (+1.0 or -1.0).

supportInserted

boolean supportInserted

Indicates if the support vector has been inserted into the map of support vectors or not. This allows us to keep the supports map to only contain the entries whose weights are non-zero.

selfKernel

double selfKernel

This is the value of the kernel applied to the example and itself. We use this value in the update step, so we can cache it for a performance boost.

previousStepWeight

double previousStepWeight

The weight of the entry on the previous step. This is used at the end of the step to calculate the total change of weights in the step.

Package gov.sandia.cognition.learning.algorithm.tree

Class gov.sandia.cognition.learning.algorithm.tree.AbstractDecisionTreeLearner extends AbstractIterativeAlgorithm implements Serializable

Serialized Fields

deciderLearner

DeciderLearner<InputType,OutputType,CategoryType,DeciderType extends Categorizer<? super InputType,? extends CategoryType>> deciderLearner

The learning algorithm for the decision function.

Class gov.sandia.cognition.learning.algorithm.tree.AbstractDecisionTreeNode extends Object implements Serializable

Serialized Fields

parent

DecisionTreeNode<InputType,OutputType> parent

The parent node of this node.

childMap

Map<K,V> childMap

The mapping of decider decision values to child nodes. For a leaf node, this can be null or empty.

decider

Categorizer<InputType,CategoryType> decider

The decider used to make a decision as to which child use. For a leaf node, the decider should be null.

incomingValue

Object incomingValue

The incoming value for the node. Usually if this is null it means that the node is a root node.

Class gov.sandia.cognition.learning.algorithm.tree.AbstractVectorThresholdMaximumGainLearner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

dimensionsToConsider

int[] dimensionsToConsider

The array of dimensions for the learner to consider. If this is null, then all dimensions are considered.

Class gov.sandia.cognition.learning.algorithm.tree.CategorizationTree extends DecisionTree<InputType,OutputType> implements Serializable

Serialized Fields

categories

Set<E> categories

The list of possible output categories.

Class gov.sandia.cognition.learning.algorithm.tree.CategorizationTreeLearner extends AbstractDecisionTreeLearner<InputType,OutputType> implements Serializable

Serialized Fields

leafCountThreshold

int leafCountThreshold

The threshold for making a node a leaf, determined by how many instances fall in the threshold.

maxDepth

int maxDepth

The maximum depth for the tree. Ignored if less than 1.

priors

Map<K,V> priors

Prior probabilities for the different categories. If null, the priors default to the category frequencies of the training data.

trainCounts

Map<K,V> trainCounts

How often each category appears in training data.

Class gov.sandia.cognition.learning.algorithm.tree.CategorizationTreeNode extends AbstractDecisionTreeNode<InputType,OutputType,InteriorType> implements Serializable

Serialized Fields

outputCategory

Object outputCategory

The output category of the node. All nodes should have this, but it is absolutely required for a leaf node.

Class gov.sandia.cognition.learning.algorithm.tree.DecisionTree extends AbstractCloneableSerializable implements Serializable

Serialized Fields

rootNode

DecisionTreeNode<InputType,OutputType> rootNode

The root node of the decision tree.

Class gov.sandia.cognition.learning.algorithm.tree.RandomSubVectorThresholdLearner extends AbstractRandomized implements Serializable

Serialized Fields

subLearner

DeciderLearner<InputType,OutputType,CategoryType,DeciderType extends Categorizer<? super InputType,? extends CategoryType>> subLearner

The decider learner for the subspace.

percentToSample

double percentToSample

The percentage of the dimensionality to sample.

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The vector factory to use.

Class gov.sandia.cognition.learning.algorithm.tree.RegressionTree extends DecisionTree<InputType,Double> implements Serializable

Class gov.sandia.cognition.learning.algorithm.tree.RegressionTreeLearner extends AbstractDecisionTreeLearner<InputType,Double> implements Serializable

Serialized Fields

regressionLearner

BatchLearner<DataType,ResultType> regressionLearner

The learning algorithm for the regression function.

leafCountThreshold

int leafCountThreshold

The threshold for making a node a leaf, determined by how many instances fall in the threshold.

maxDepth

int maxDepth

The maximum depth for the tree. Ignored if less than 1.

Class gov.sandia.cognition.learning.algorithm.tree.RegressionTreeNode extends AbstractDecisionTreeNode<InputType,Double,InteriorType> implements Serializable

Serialized Fields

scalarFunction

Evaluator<InputType,OutputType> scalarFunction

The function to apply for leaf nodes.

value

double value

The value stored at the tree node. It is used as a backup value if no scalar function exists for the node but the output is requested.

Class gov.sandia.cognition.learning.algorithm.tree.VectorThresholdGiniImpurityLearner extends AbstractVectorThresholdMaximumGainLearner<OutputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.tree.VectorThresholdHellingerDistanceLearner extends AbstractVectorThresholdMaximumGainLearner<OutputType> implements Serializable

Class gov.sandia.cognition.learning.algorithm.tree.VectorThresholdInformationGainLearner extends AbstractVectorThresholdMaximumGainLearner<OutputType> implements Serializable

Serialized Fields

categories

ArrayList<E> categories

categoryPriors

double[] categoryPriors

categoryCounts

int[] categoryCounts

categoryProbabilities

double[] categoryProbabilities

Following is scratch space used when computing weighted entropy. It is declared here so it can be allocated once, instead of during every entropy evaluation.

Class gov.sandia.cognition.learning.algorithm.tree.VectorThresholdVarianceLearner extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.data

Class gov.sandia.cognition.learning.data.AbstractInputOutputPair extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.data.AbstractTargetEstimatePair extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.data.AbstractValueDiscriminantPair extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.data.DefaultInputOutputPair extends AbstractInputOutputPair<InputType,OutputType> implements Serializable

Serialized Fields

input

Object input

The input.

output

Object output

The output associated with the input.

Class gov.sandia.cognition.learning.data.DefaultTargetEstimatePair extends AbstractTargetEstimatePair<TargetType,EstimateType> implements Serializable

Serialized Fields

target

Object target

Target (ground-truth) value.

estimate

Object estimate

Estimate (prediction) of the target value.

Class gov.sandia.cognition.learning.data.DefaultValueDiscriminantPair extends AbstractValueDiscriminantPair<ValueType,DiscriminantType extends Comparable<? super DiscriminantType>> implements Serializable

Serialized Fields

value

Object value

The value.

discriminant

Comparable<T> discriminant

The discriminant.

Class gov.sandia.cognition.learning.data.DefaultWeightedInputOutputPair extends DefaultInputOutputPair<InputType,OutputType> implements Serializable

Serialized Fields

weight

double weight

Weighting term for the InputOutputPair.

Class gov.sandia.cognition.learning.data.DefaultWeightedTargetEstimatePair extends DefaultTargetEstimatePair<TargetType,EstimateType> implements Serializable

Serialized Fields

weight

double weight

The weight.

Class gov.sandia.cognition.learning.data.DefaultWeightedValueDiscriminant extends DefaultWeightedValue<ValueType> implements Serializable

Class gov.sandia.cognition.learning.data.RandomDataPartitioner extends AbstractRandomized implements Serializable

Serialized Fields

trainingPercent

double trainingPercent

The percentage of training data.

Package gov.sandia.cognition.learning.data.feature

Class gov.sandia.cognition.learning.data.feature.DelayFunction extends AbstractStatefulEvaluator<DataType,DataType,FiniteCapacityBuffer<DataType>> implements Serializable

Serialized Fields

delaySamples

int delaySamples

Number of samples to delay the value

Class gov.sandia.cognition.learning.data.feature.LinearRegressionCoefficientExtractor extends AbstractStatefulEvaluator<Vector,Vector,FiniteCapacityBuffer<Vector>> implements Serializable

Serialized Fields

maxBufferSize

int maxBufferSize

maximum number of vectors to hold in the buffer

Class gov.sandia.cognition.learning.data.feature.MultivariateDecorrelator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

gaussian

MultivariateGaussian gaussian

The underlying Gaussian.

covarianceInverseSquareRoot

Matrix covarianceInverseSquareRoot

The square root of the inverse of the covariance.

Class gov.sandia.cognition.learning.data.feature.MultivariateDecorrelator.DiagonalCovarianceLearner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

defaultCovariance

double defaultCovariance

The default covariance. Added to the diagonal to prevent it from becoming singular.

Class gov.sandia.cognition.learning.data.feature.MultivariateDecorrelator.FullCovarianceLearner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

defaultCovariance

double defaultCovariance

The default covariance. Added to the diagonal to prevent it from becoming singular.

Class gov.sandia.cognition.learning.data.feature.RandomSubspace extends AbstractRandomized implements Serializable

Serialized Fields

size

int size

The size of the random subspace to create, which is the number of dimensions that are chosen.

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The vector factory for the sub vector evaluator to use.

Class gov.sandia.cognition.learning.data.feature.StandardDistributionNormalizer extends AbstractUnivariateScalarFunction implements Serializable

Serialized Fields

mean

double mean

The mean of normalization.

variance

double variance

The variance for normalization.

standardDeviation

double standardDeviation

The cached value of the standard deviation for normalization.

Class gov.sandia.cognition.learning.data.feature.StandardDistributionNormalizer.Learner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

outlierPercent

double outlierPercent

The percentage of outliers to exclude from learning.

Package gov.sandia.cognition.learning.experiment

Class gov.sandia.cognition.learning.experiment.AbstractLearningExperiment extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.experiment.AbstractValidationFoldExperiment extends AbstractLearningExperiment implements Serializable

Serialized Fields

foldCreator

ValidationFoldCreator<InputDataType,FoldDataType> foldCreator

The method to use to create the validation folds.

numTrials

int numTrials

The number of trials in the experiment, which is the number of folds in the experiment.

Class gov.sandia.cognition.learning.experiment.CrossFoldCreator extends AbstractRandomized implements Serializable

Serialized Fields

numFolds

int numFolds

The number of folds to create.

Class gov.sandia.cognition.learning.experiment.LearnerComparisonExperiment extends AbstractValidationFoldExperiment<InputDataType,FoldDataType> implements Serializable

Serialized Fields

performanceEvaluator

PerformanceEvaluator<ObjectType,DataType,ResultType> performanceEvaluator

The evaluator to use to compute the performance of the learned object on each fold.

summarizer

Summarizer<DataType,SummaryType> summarizer

The summarizer for summarizing the result of the performance evaluator from all the folds.

statisticalTest

NullHypothesisEvaluator<DataType> statisticalTest

The statistical test to use to determine if the two learners are significantly different.

learners

Pair<FirstType,SecondType> learners

The learners that the experiment is being performed on.

statistics

DefaultPair<FirstType,SecondType> statistics

The performance evaluations made during the experiment.

confidence

ConfidenceStatistic confidence

The confidence statistic generated from the underlying performance statistics.

summaries

DefaultPair<FirstType,SecondType> summaries

The summaries of performance.

Class gov.sandia.cognition.learning.experiment.LearnerRepeatExperiment extends AbstractLearningExperiment implements Serializable

Serialized Fields

numTrials

int numTrials

The number of trials to repeat the learning.

performanceEvaluator

PerformanceEvaluator<ObjectType,DataType,ResultType> performanceEvaluator

The evaluator to use to compute the performance of the learned object on each fold.

summarizer

Summarizer<DataType,SummaryType> summarizer

The summarizer for summarizing the result of the performance evaluator from all the folds.

learner

BatchLearner<DataType,ResultType> learner

The learner that the experiment is run on.

statistics

ArrayList<E> statistics

The performance evaluations made during the experiment.

summary

Object summary

The summary of the performance evaluations made at the end of the experiment.

Class gov.sandia.cognition.learning.experiment.LearnerValidationExperiment extends AbstractValidationFoldExperiment<InputDataType,FoldDataType> implements Serializable

Serialized Fields

performanceEvaluator

PerformanceEvaluator<ObjectType,DataType,ResultType> performanceEvaluator

The evaluator to use to compute the performance of the learned object on each fold.

summarizer

Summarizer<DataType,SummaryType> summarizer

The summarizer for summarizing the result of the performance evaluator from all the folds.

learner

BatchLearner<DataType,ResultType> learner

The learner that the experiment is run on.

statistics

ArrayList<E> statistics

The performance evaluations made during the experiment.

summary

Object summary

The summary of the performance evaluations made at the end of the experiment.

Class gov.sandia.cognition.learning.experiment.LeaveOneOutFoldCreator extends Object implements Serializable

Class gov.sandia.cognition.learning.experiment.OnlineLearnerValidationExperiment extends AbstractLearningExperiment implements Serializable

Serialized Fields

performanceEvaluator

PerformanceEvaluator<ObjectType,DataType,ResultType> performanceEvaluator

The evaluator to use to compute the performance of the learned object on each fold.

summarizer

Summarizer<DataType,SummaryType> summarizer

The summarizer for summarizing the result of the performance evaluator from all the folds.

numTrials

int numTrials

The number of trials in the experiment, which is the number of folds in the experiment.

statistics

ArrayList<E> statistics

The performance evaluations made during the experiment.

summary

Object summary

The summary of the performance evaluations made at the end of the experiment.

Class gov.sandia.cognition.learning.experiment.ParallelLearnerValidationExperiment extends LearnerValidationExperiment<InputDataType,FoldDataType,LearnedType,StatisticType,SummaryType> implements Serializable

Class gov.sandia.cognition.learning.experiment.RandomByTwoFoldCreator extends AbstractRandomized implements Serializable

Serialized Fields

numSplits

int numSplits

The number of splits. The number of folds is twice this number.

Class gov.sandia.cognition.learning.experiment.RandomFoldCreator extends Object implements Serializable

Serialized Fields

numFolds

int numFolds

The number of folds to create.

partitioner

RandomizedDataPartitioner<DataType> partitioner

The partitioner used for each fold.

Class gov.sandia.cognition.learning.experiment.SupervisedLearnerComparisonExperiment extends LearnerComparisonExperiment<InputOutputPair<InputType,OutputType>,InputOutputPair<InputType,OutputType>,Evaluator<? super InputType,OutputType>,StatisticType,SummaryType> implements Serializable

Class gov.sandia.cognition.learning.experiment.SupervisedLearnerValidationExperiment extends LearnerValidationExperiment<InputOutputPair<InputType,OutputType>,InputOutputPair<InputType,OutputType>,Evaluator<? super InputType,OutputType>,StatisticType,SummaryType> implements Serializable

Package gov.sandia.cognition.learning.function

Class gov.sandia.cognition.learning.function.ConstantEvaluator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

value

Object value

The output value.

Class gov.sandia.cognition.learning.function.LinearCombinationFunction extends AbstractCloneableSerializable implements Serializable

Serialized Fields

basisFunctions

ArrayList<E> basisFunctions

Collection of basis functions to combine to produce the output

coefficients

Vector coefficients

Coefficients for the basisFunctions

Package gov.sandia.cognition.learning.function.categorization

Class gov.sandia.cognition.learning.function.categorization.AbstractBinaryCategorizer extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.categorization.AbstractCategorizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

categories

Set<E> categories

The set of categories that are the possible output values of the categorizer.

Class gov.sandia.cognition.learning.function.categorization.AbstractConfidenceWeightedBinaryCategorizer extends LinearBinaryCategorizer implements Serializable

Class gov.sandia.cognition.learning.function.categorization.AbstractDiscriminantBinaryCategorizer extends AbstractBinaryCategorizer<InputType> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.AbstractDiscriminantCategorizer extends AbstractCategorizer<InputType,CategoryType> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.AbstractThresholdBinaryCategorizer extends AbstractDiscriminantBinaryCategorizer<InputType> implements Serializable

Serialized Fields

threshold

double threshold

Threshold, below which I will return lowValue, above or equal to I will return highValue.

Class gov.sandia.cognition.learning.function.categorization.BinaryVersusCategorizer extends AbstractDiscriminantCategorizer<InputType,CategoryType,Double> implements Serializable

Serialized Fields

categoryPairsToEvaluatorMap

Map<K,V> categoryPairsToEvaluatorMap

Maps false-true category pairs .

Class gov.sandia.cognition.learning.function.categorization.BinaryVersusCategorizer.Learner extends AbstractBatchLearnerContainer<BatchLearner<? super Collection<? extends InputOutputPair<? extends InputType,Boolean>>,? extends Evaluator<? super InputType,Boolean>>> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.CompositeCategorizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

preprocessor

Evaluator<InputType,OutputType> preprocessor

The preprocessor for the input data.

categorizer

Categorizer<InputType,CategoryType> categorizer

The categorizer.

Class gov.sandia.cognition.learning.function.categorization.DefaultConfidenceWeightedBinaryCategorizer extends AbstractConfidenceWeightedBinaryCategorizer implements Serializable

Serialized Fields

covariance

Matrix covariance

The covariance matrix.

Class gov.sandia.cognition.learning.function.categorization.DefaultKernelBinaryCategorizer extends KernelBinaryCategorizer<InputType,DefaultWeightedValue<InputType>> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.DiagonalConfidenceWeightedBinaryCategorizer extends AbstractConfidenceWeightedBinaryCategorizer implements Serializable

Serialized Fields

variance

Vector variance

The variance values, which is the diagonal of the covariance matrix. It is stored as a vector to avoid needing to instantiate and use matrix operations with it.

Class gov.sandia.cognition.learning.function.categorization.EvaluatorToCategorizerAdapter extends AbstractCategorizer<InputType,CategoryType> implements Serializable

Serialized Fields

evaluator

Evaluator<InputType,OutputType> evaluator

The evaluator that is being wrapped into a categorizer.

Class gov.sandia.cognition.learning.function.categorization.EvaluatorToCategorizerAdapter.Learner extends AbstractBatchLearnerContainer<BatchLearner<? super Collection<? extends InputOutputPair<? extends InputType,CategoryType>>,? extends Evaluator<? super InputType,? extends CategoryType>>> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.FisherLinearDiscriminantBinaryCategorizer extends ScalarFunctionToBinaryCategorizerAdapter<Vector> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.FisherLinearDiscriminantBinaryCategorizer.ClosedFormSolver extends AbstractCloneableSerializable implements Serializable

Serialized Fields

defaultCovariance

double defaultCovariance

The default covariance.

Class gov.sandia.cognition.learning.function.categorization.KernelBinaryCategorizer extends AbstractDiscriminantBinaryCategorizer<InputType> implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The internal kernel.

examples

Collection<E> examples

The list of weighted examples that are used for categorization.

bias

double bias

The bias term.

Class gov.sandia.cognition.learning.function.categorization.LinearBinaryCategorizer extends AbstractDiscriminantBinaryCategorizer<Vectorizable> implements Serializable

Serialized Fields

weights

Vector weights

The weight vector.

bias

double bias

The bias term.

Class gov.sandia.cognition.learning.function.categorization.LinearMultiCategorizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

prototypes

Map<K,V> prototypes

A map of each category to its prototype categorizer.

Class gov.sandia.cognition.learning.function.categorization.MaximumAPosterioriCategorizer extends AbstractDistribution<ObservationType> implements Serializable

Serialized Fields

categoryPriors

DataDistribution.PMF<KeyType> categoryPriors

PMF of the various categories

categoryConditionals

Map<K,V> categoryConditionals

Map that contains the probability functions for the observations for the given categories.

Class gov.sandia.cognition.learning.function.categorization.MaximumAPosterioriCategorizer.Learner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

conditionalLearner

BatchLearner<DataType,ResultType> conditionalLearner

Learner that creates the conditional distributions for each category.

Class gov.sandia.cognition.learning.function.categorization.ScalarFunctionToBinaryCategorizerAdapter extends AbstractThresholdBinaryCategorizer<InputType> implements Serializable

Serialized Fields

evaluator

Evaluator<InputType,OutputType> evaluator

The scalar evaluator.

Class gov.sandia.cognition.learning.function.categorization.ScalarThresholdBinaryCategorizer extends AbstractThresholdBinaryCategorizer<Double> implements Serializable

Class gov.sandia.cognition.learning.function.categorization.VectorElementThresholdCategorizer extends AbstractThresholdBinaryCategorizer<Vectorizable> implements Serializable

Serialized Fields

index

int index

The index to apply the threshold to.

Class gov.sandia.cognition.learning.function.categorization.WinnerTakeAllCategorizer extends AbstractDiscriminantCategorizer<InputType,CategoryType,Double> implements Serializable

Serialized Fields

evaluator

Evaluator<InputType,OutputType> evaluator

The evaluator that outputs a vector to return.

Class gov.sandia.cognition.learning.function.categorization.WinnerTakeAllCategorizer.Learner extends AbstractBatchLearnerContainer<BatchLearner<? super Collection<? extends InputOutputPair<? extends InputType,Vector>>,? extends Evaluator<? super InputType,? extends Vectorizable>>> implements Serializable

Serialized Fields

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The vector factory used.

Package gov.sandia.cognition.learning.function.cost

Class gov.sandia.cognition.learning.function.cost.AbstractCostFunction extends AbstractCloneableSerializable implements Serializable

Serialized Fields

costParameters

Object costParameters

The parameters of the cost function.

Class gov.sandia.cognition.learning.function.cost.AbstractParallelizableCostFunction extends AbstractSupervisedCostFunction<Vector,Vector> implements Serializable

Class gov.sandia.cognition.learning.function.cost.AbstractSupervisedCostFunction extends AbstractSupervisedPerformanceEvaluator<InputType,TargetType,TargetType,Double> implements Serializable

Serialized Fields

costParameters

Collection<E> costParameters

Labeled dataset to use to evaluate the cost against

Class gov.sandia.cognition.learning.function.cost.ClusterDistortionMeasure extends AbstractCloneableSerializable implements Serializable

Serialized Fields

costParameters

ClusterDivergenceFunction<ClusterType extends Cluster<DataType>,DataType> costParameters

Divergence function that defines the cost function

Class gov.sandia.cognition.learning.function.cost.EuclideanDistanceCostFunction extends AbstractCloneableSerializable implements Serializable

Serialized Fields

goal

Vector goal

The goal of the cost function.

Class gov.sandia.cognition.learning.function.cost.KolmogorovSmirnovDivergence extends AbstractCostFunction<UnivariateDistribution<DataType extends Number>,Collection<? extends DataType extends Number>> implements Serializable

Class gov.sandia.cognition.learning.function.cost.MeanL1CostFunction extends AbstractSupervisedCostFunction<Vector,Vector> implements Serializable

Class gov.sandia.cognition.learning.function.cost.MeanSquaredErrorCostFunction extends AbstractSupervisedCostFunction<Vector,Vector> implements Serializable

Class gov.sandia.cognition.learning.function.cost.NegativeLogLikelihood extends AbstractCostFunction<ComputableDistribution<DataType>,Collection<? extends DataType>> implements Serializable

Class gov.sandia.cognition.learning.function.cost.ParallelClusterDistortionMeasure extends ClusterDistortionMeasure<DataType,ClusterType extends Cluster<DataType>> implements Serializable

Class gov.sandia.cognition.learning.function.cost.ParallelizedCostFunctionContainer extends AbstractSupervisedCostFunction<Vector,Vector> implements Serializable

Serialized Fields

costFunction

ParallelizableCostFunction costFunction

Cost function to parallelize

Class gov.sandia.cognition.learning.function.cost.ParallelNegativeLogLikelihood extends NegativeLogLikelihood<DataType> implements Serializable

Class gov.sandia.cognition.learning.function.cost.SumSquaredErrorCostFunction extends AbstractParallelizableCostFunction implements Serializable

Class gov.sandia.cognition.learning.function.cost.SumSquaredErrorCostFunction.Cache extends AbstractCloneableSerializable implements Serializable

Serialized Fields

J

Matrix J

Jacobian

JtJ

Matrix JtJ

Inner-product of the Jacobian matrix: J.transpose().times( J )

Jte

Vector Jte

Jacobian transpose times Error: J.transpose().times( error )

parameterCost

double parameterCost

Cost-function value of the parameter set

Class gov.sandia.cognition.learning.function.cost.SumSquaredErrorCostFunction.GradientPartialSSE extends DefaultPair<Vector,Double> implements Serializable

Package gov.sandia.cognition.learning.function.distance

Class gov.sandia.cognition.learning.function.distance.ChebyshevDistanceMetric extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.distance.CosineDistanceMetric extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.distance.DefaultDivergenceFunctionContainer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

divergenceFunction

DivergenceFunction<FirstType,SecondType> divergenceFunction

The internal divergence function for the object to use.

Class gov.sandia.cognition.learning.function.distance.DivergencesEvaluator extends DefaultVectorFactoryContainer implements Serializable

Serialized Fields

divergenceFunction

DivergenceFunction<FirstType,SecondType> divergenceFunction

The divergence function to apply between the data and the input.

values

Collection<E> values

The data to evaluate the divergence from.

Class gov.sandia.cognition.learning.function.distance.DivergencesEvaluator.Learner extends AbstractBatchLearnerContainer<BatchLearner<? super DataType,? extends Collection<ValueType>>> implements Serializable

Serialized Fields

divergenceFunction

DivergenceFunction<FirstType,SecondType> divergenceFunction

The divergence function to apply between the data and the input.

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The vector factory to use.

Class gov.sandia.cognition.learning.function.distance.EuclideanDistanceMetric extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.distance.EuclideanDistanceSquaredMetric extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.distance.IdentityDistanceMetric extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.distance.ManhattanDistanceMetric extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.distance.MinkowskiDistanceMetric extends AbstractCloneableSerializable implements Serializable

Serialized Fields

power

double power

The power that the distance is computed to.

Class gov.sandia.cognition.learning.function.distance.WeightedEuclideanDistanceMetric extends AbstractCloneableSerializable implements Serializable

Serialized Fields

weights

Vector weights

The weights assigned to each dimension for the distance. The weights cannot be negative.

Package gov.sandia.cognition.learning.function.kernel

Class gov.sandia.cognition.learning.function.kernel.DefaultKernelContainer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

kernel

Kernel<InputType> kernel

The internal kernel.

Class gov.sandia.cognition.learning.function.kernel.DefaultKernelsContainer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

kernels

Collection<E> kernels

The collection of kernels in the container.

Class gov.sandia.cognition.learning.function.kernel.ExponentialKernel extends DefaultKernelContainer<InputType> implements Serializable

Class gov.sandia.cognition.learning.function.kernel.KernelDistanceMetric extends DefaultKernelContainer<InputType> implements Serializable

Class gov.sandia.cognition.learning.function.kernel.LinearKernel extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.kernel.NormalizedKernel extends DefaultKernelContainer<InputType> implements Serializable

Class gov.sandia.cognition.learning.function.kernel.PolynomialKernel extends AbstractCloneableSerializable implements Serializable

Serialized Fields

degree

int degree

The degree of the polynomial. Must be positive.

constant

double constant

The constant for the polynomial.

Class gov.sandia.cognition.learning.function.kernel.ProductKernel extends DefaultKernelsContainer<InputType> implements Serializable

Class gov.sandia.cognition.learning.function.kernel.RadialBasisKernel extends AbstractCloneableSerializable implements Serializable

Serialized Fields

sigma

double sigma

Sigma is the parameter that controls the bandwidth of the radial basis kernel.

negativeTwoSigmaSquared

double negativeTwoSigmaSquared

This value is cached so it doesn't have to be computed each time. It is -2.0 * sigma * sigma.

Class gov.sandia.cognition.learning.function.kernel.ScalarFunctionKernel extends AbstractCloneableSerializable implements Serializable

Serialized Fields

function

Evaluator<InputType,OutputType> function

The scalar function for the kernel to use.

Class gov.sandia.cognition.learning.function.kernel.SigmoidKernel extends AbstractCloneableSerializable implements Serializable

Serialized Fields

kappa

double kappa

The kappa value to multiply times the dot product.

constant

double constant

The constant used in the sigmoid.

Class gov.sandia.cognition.learning.function.kernel.SumKernel extends DefaultKernelsContainer<InputType> implements Serializable

Class gov.sandia.cognition.learning.function.kernel.VectorFunctionKernel extends DefaultKernelContainer<Vector> implements Serializable

Serialized Fields

function

VectorFunction function

The vector function to use.

Class gov.sandia.cognition.learning.function.kernel.WeightedKernel extends DefaultKernelContainer<InputType> implements Serializable

Serialized Fields

weight

double weight

The weight on the kernel. Must be non-negative.

Class gov.sandia.cognition.learning.function.kernel.ZeroKernel extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.function.regression

Class gov.sandia.cognition.learning.function.regression.AbstractRegressor extends AbstractScalarFunction<InputType> implements Serializable

Package gov.sandia.cognition.learning.function.scalar

Class gov.sandia.cognition.learning.function.scalar.AtanFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Serialized Fields

scaleFactor

double scaleFactor

Scales the Math.atan() value to ensure that it has the desired "maxMagnitude"

Class gov.sandia.cognition.learning.function.scalar.CosineFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Serialized Fields

amplitude

double amplitude

Maximum value of the cosine function. The value of "A" in y=A*cos(2*pi*f*x + phase).

frequency

double frequency

Frequency of the cosine function. The value of "f" in y=A*cos(2*pi*f*x + phase).

phase

double phase

Phase of the cosine function. The value of "phase" in y=A*cos(2*pi*f*x + phase).

Class gov.sandia.cognition.learning.function.scalar.IdentityScalarFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Class gov.sandia.cognition.learning.function.scalar.KernelScalarFunction extends DefaultKernelContainer<InputType> implements Serializable

Serialized Fields

examples

Collection<E> examples

The list of weighted examples that are used for categorization.

bias

double bias

The bias term.

Class gov.sandia.cognition.learning.function.scalar.KolmogorovSmirnovEvaluator extends AbstractStatefulEvaluator<Double,Double,FiniteCapacityBuffer<Double>> implements Serializable

Serialized Fields

cdf

CumulativeDistributionFunction<NumberType extends Number> cdf

The cumulative distribution function to base the evaluator on.

capacity

int capacity

The capacity of the state.

Class gov.sandia.cognition.learning.function.scalar.LinearCombinationScalarFunction extends LinearCombinationFunction<InputType,Double> implements Serializable

Class gov.sandia.cognition.learning.function.scalar.LinearDiscriminant extends AbstractRegressor<Vectorizable> implements Serializable

Serialized Fields

weightVector

Vector weightVector

Weight Vector to dot-product with the input

Class gov.sandia.cognition.learning.function.scalar.LinearDiscriminantWithBias extends LinearDiscriminant implements Serializable

Serialized Fields

bias

double bias

Bias term that gets added to the output of the dot product.

Class gov.sandia.cognition.learning.function.scalar.LinearFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Serialized Fields

slope

double slope

The slope (m).

offset

double offset

The offset (b).

Class gov.sandia.cognition.learning.function.scalar.LinearVectorScalarFunction extends AbstractRegressor<Vectorizable> implements Serializable

Serialized Fields

weights

Vector weights

The weight vector.

bias

double bias

The bias term.

Class gov.sandia.cognition.learning.function.scalar.LocallyWeightedKernelScalarFunction extends KernelScalarFunction<InputType> implements Serializable

Serialized Fields

constantWeight

double constantWeight

The constant weight is used as a weight for the constant value that is added to the result to bias the function to the constant value.

constantValue

double constantValue

The constant value is what the constant weight biases the function toward when there is near zero weight.

Class gov.sandia.cognition.learning.function.scalar.PolynomialFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Serialized Fields

exponent

double exponent

Real-valued exponent of this polynomial

Class gov.sandia.cognition.learning.function.scalar.PolynomialFunction.Cubic extends PolynomialFunction.Quadratic implements Serializable

Serialized Fields

q3

double q3

Cubic (third-order) coefficient

Class gov.sandia.cognition.learning.function.scalar.PolynomialFunction.Linear extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Serialized Fields

q0

double q0

Constant (zeroth-order) coefficient

q1

double q1

Linear (first-order) coefficient

Class gov.sandia.cognition.learning.function.scalar.PolynomialFunction.Quadratic extends PolynomialFunction.Linear implements Serializable

Serialized Fields

q2

double q2

Quadratic (second-order) coefficient

Class gov.sandia.cognition.learning.function.scalar.PolynomialFunction.Regression extends AbstractCloneableSerializable implements Serializable

Serialized Fields

polynomials

ScalarBasisSet<InputType> polynomials

Polynomials to use in the regression

Class gov.sandia.cognition.learning.function.scalar.SigmoidFunction extends AbstractDifferentiableUnivariateScalarFunction implements Serializable

Class gov.sandia.cognition.learning.function.scalar.ThresholdFunction extends AbstractUnivariateScalarFunction implements Serializable

Serialized Fields

highValue

double highValue

Values above (or equal to) threshold are assigned this value

lowValue

double lowValue

Values below threshold are assigned this value

threshold

double threshold

Current threshold, below which a value is assigned lowValue, above which (or equal to) is assigned highValue

Class gov.sandia.cognition.learning.function.scalar.VectorEntryFunction extends AbstractScalarFunction<Vectorizable> implements Serializable

Serialized Fields

index

int index

The index of the vector to get.

Class gov.sandia.cognition.learning.function.scalar.VectorFunctionLinearDiscriminant extends AbstractRegressor<InputType> implements Serializable

Serialized Fields

vectorFunction

Evaluator<InputType,OutputType> vectorFunction

Maps the input space to a Vector

discriminant

LinearDiscriminant discriminant

The dot product of the discriminant with the output of the vectorFunction is the output (scalar) value. Must have the same dimensions as the outputDimensionality of vectorFunction.

Class gov.sandia.cognition.learning.function.scalar.VectorFunctionToScalarFunction extends AbstractRegressor<InputType> implements Serializable

Serialized Fields

vectorFunction

Evaluator<InputType,OutputType> vectorFunction

The function that takes a given input and outputs a 1-dimensional vector.

Class gov.sandia.cognition.learning.function.scalar.VectorFunctionToScalarFunction.Learner extends AbstractCloneableSerializable implements Serializable

Serialized Fields

vectorLearner

BatchLearner<DataType,ResultType> vectorLearner

The supervised learner that learns on vectors as outputs.

Package gov.sandia.cognition.learning.function.summarizer

Class gov.sandia.cognition.learning.function.summarizer.MostFrequentSummarizer extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.learning.function.vector

Class gov.sandia.cognition.learning.function.vector.DifferentiableFeedforwardNeuralNetwork extends FeedforwardNeuralNetwork implements Serializable

Class gov.sandia.cognition.learning.function.vector.DifferentiableGeneralizedLinearModel extends GeneralizedLinearModel implements Serializable

Class gov.sandia.cognition.learning.function.vector.ElementWiseDifferentiableVectorFunction extends ElementWiseVectorFunction implements Serializable

Class gov.sandia.cognition.learning.function.vector.ElementWiseVectorFunction extends AbstractCloneableSerializable implements Serializable

Serialized Fields

scalarFunction

UnivariateScalarFunction scalarFunction

Underlying scalar function to apply to each Vector element

Class gov.sandia.cognition.learning.function.vector.EntropyEvaluator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.function.vector.FeedforwardNeuralNetwork extends AbstractCloneableSerializable implements Serializable

Serialized Fields

layers

ArrayList<E> layers

Layers that comprise this neural network

Class gov.sandia.cognition.learning.function.vector.GaussianContextRecognizer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

gaussianMixture

MixtureOfGaussians.PDF gaussianMixture

Underlying MixtureOfGaussians that computes context probabilities

Class gov.sandia.cognition.learning.function.vector.GaussianContextRecognizer.Learner extends AnytimeAlgorithmWrapper<GaussianContextRecognizer,AnytimeBatchLearner<Collection<? extends Vector>,Collection<GaussianCluster>>> implements Serializable

Class gov.sandia.cognition.learning.function.vector.GeneralizedLinearModel extends AbstractCloneableSerializable implements Serializable

Serialized Fields

discriminant

MultivariateDiscriminant discriminant

GradientDescendable that multiplies an input by the internal matrix

squashingFunction

VectorFunction squashingFunction

VectorFunction that is applied to the output of the matrix multiply

Class gov.sandia.cognition.learning.function.vector.LinearCombinationVectorFunction extends LinearCombinationFunction<Vector,Vector> implements Serializable

Class gov.sandia.cognition.learning.function.vector.LinearVectorFunction extends AbstractCloneableSerializable implements Serializable

Serialized Fields

scaleFactor

double scaleFactor

Scales the input by this amount.

Class gov.sandia.cognition.learning.function.vector.MultivariateDiscriminant extends AbstractCloneableSerializable implements Serializable

Serialized Fields

discriminant

Matrix discriminant

Internal matrix to premultiply input vectors by.

Class gov.sandia.cognition.learning.function.vector.MultivariateDiscriminantWithBias extends MultivariateDiscriminant implements Serializable

Serialized Fields

bias

Vector bias

Bias term that gets added the output of the matrix multiplication.

Class gov.sandia.cognition.learning.function.vector.ScalarBasisSet extends AbstractCloneableSerializable implements Serializable

Serialized Fields

basisFunctions

Collection<E> basisFunctions

Collection of scalar basis functions, where the ith function operates on the ith element of the output Vector

Class gov.sandia.cognition.learning.function.vector.SubVectorEvaluator extends DefaultVectorFactoryContainer implements Serializable

Serialized Fields

inputDimensionality

int inputDimensionality

The expected dimensionality of the input. All the subIndices must be less than this value.

subIndices

int[] subIndices

The indices to pull out of an input vector to create a new vector from.

Class gov.sandia.cognition.learning.function.vector.ThreeLayerFeedforwardNeuralNetwork extends AbstractRandomized implements Serializable

Serialized Fields

inputToHiddenWeights

Matrix inputToHiddenWeights

Matrix of weights to pre-multiply the inputs by.

inputToHiddenBiasWeights

Vector inputToHiddenBiasWeights

Bias weights to add to each of the hidden units.

hiddenToOutputWeights

Matrix hiddenToOutputWeights

Matrix of weights to pre-multiply the hidden-unit activations by.

hiddenToOutputBiasWeights

Vector hiddenToOutputBiasWeights

Bias weights to add to each of the output units.

squashingFunction

DifferentiableUnivariateScalarFunction squashingFunction

Squashing function to apply at the hidden layer.

initializationRange

double initializationRange

Range of values to initialize the weights between, must be greater than or equal to zero.

Class gov.sandia.cognition.learning.function.vector.VectorizableVectorConverter extends Object implements Serializable

Class gov.sandia.cognition.learning.function.vector.VectorizableVectorConverterWithBias extends VectorizableVectorConverter implements Serializable

Serialized Fields

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The factory used to create the vector.

Package gov.sandia.cognition.learning.parameter

Class gov.sandia.cognition.learning.parameter.ParameterAdaptableBatchLearnerWrapper extends AbstractBatchLearnerContainer<LearnerType extends BatchLearner<? super DataType,? extends ResultType>> implements Serializable

Serialized Fields

parameterAdapters

LinkedList<E> parameterAdapters

The list of parameter adapters for the learner. It should be null if there are no adapters.

Package gov.sandia.cognition.learning.performance

Class gov.sandia.cognition.learning.performance.AbstractSupervisedPerformanceEvaluator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.performance.MeanAbsoluteErrorEvaluator extends AbstractSupervisedPerformanceEvaluator<InputType,Double,Double,Double> implements Serializable

Class gov.sandia.cognition.learning.performance.MeanSquaredErrorEvaluator extends AbstractSupervisedPerformanceEvaluator<InputType,Double,Double,Double> implements Serializable

Class gov.sandia.cognition.learning.performance.MeanZeroOneErrorEvaluator extends AbstractSupervisedCostFunction<InputType,DataType> implements Serializable

Class gov.sandia.cognition.learning.performance.RootMeanSquaredErrorEvaluator extends AbstractSupervisedPerformanceEvaluator<InputType,Double,Double,Double> implements Serializable

Package gov.sandia.cognition.learning.performance.categorization

Class gov.sandia.cognition.learning.performance.categorization.AbstractBinaryConfusionMatrix extends AbstractConfusionMatrix<Boolean> implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.AbstractConfusionMatrix extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.ConfusionMatrixPerformanceEvaluator extends AbstractSupervisedPerformanceEvaluator<InputType,CategoryType,CategoryType,ConfusionMatrix<CategoryType>> implements Serializable

Serialized Fields

factory

Factory<CreatedType> factory

The factory used to create the confusion matrix of the evaluator.

Class gov.sandia.cognition.learning.performance.categorization.DefaultBinaryConfusionMatrix extends AbstractBinaryConfusionMatrix implements Serializable

Serialized Fields

trueNegativesCount

double trueNegativesCount

Number of true negatives. The (false, false) entry.

falsePositivesCount

double falsePositivesCount

Number of false positives. The (false, true) entry.

falseNegativesCount

double falseNegativesCount

Number of false negatives. The (true, false) entry.

truePositivesCount

double truePositivesCount

Number of true positives. The (true, true) entry.

Class gov.sandia.cognition.learning.performance.categorization.DefaultBinaryConfusionMatrix.ActualPredictedPairSummarizer extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.DefaultBinaryConfusionMatrix.CombineSummarizer extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.DefaultBinaryConfusionMatrix.PerformanceEvaluator extends AbstractSupervisedPerformanceEvaluator<InputType,Boolean,Boolean,DefaultBinaryConfusionMatrix> implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.DefaultBinaryConfusionMatrixConfidenceInterval.Summary extends AbstractCloneableSerializable implements Serializable

Serialized Fields

confidence

double confidence

The confidence for the created interval.

Class gov.sandia.cognition.learning.performance.categorization.DefaultConfusionMatrix extends AbstractConfusionMatrix<CategoryType> implements Serializable

Serialized Fields

confusions

Map<K,V> confusions

The backing map of confusion matrix entries. The first key is the actual category and the second is the predicted category.

Class gov.sandia.cognition.learning.performance.categorization.DefaultConfusionMatrix.ActualPredictedPairSummarizer extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.DefaultConfusionMatrix.CombineSummarizer extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.learning.performance.categorization.DefaultConfusionMatrix.Factory extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.math

Class gov.sandia.cognition.math.AbstractDifferentiableUnivariateScalarFunction extends AbstractUnivariateScalarFunction implements Serializable

Class gov.sandia.cognition.math.AbstractEuclideanRing extends AbstractRing<RingType extends EuclideanRing<RingType>> implements Serializable

Class gov.sandia.cognition.math.AbstractField extends AbstractEuclideanRing<FieldType extends Field<FieldType>> implements Serializable

Class gov.sandia.cognition.math.AbstractRing extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.AbstractScalarFunction extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.AbstractUnivariateScalarFunction extends AbstractScalarFunction<Double> implements Serializable

Class gov.sandia.cognition.math.ComplexNumber extends AbstractField<ComplexNumber> implements Serializable

Serialized Fields

realPart

double realPart

The real part of the ComplexNumber

imaginaryPart

double imaginaryPart

The imaginary part of the ComplexNumber

Class gov.sandia.cognition.math.LentzMethod extends AbstractAnytimeAlgorithm<Double> implements Serializable

Serialized Fields

tolerance

double tolerance

Tolerance of the algorithm for convergence

minDenominator

double minDenominator

Value to keep denominators from equaling 0.0

currentC

double currentC

Current value of the "C" variable in Lentz's method

currentD

double currentD

Current value of the "D" variable in Lentz's method

result

Double result

Value of the continued fraction, null if not valid

fractionValue

double fractionValue

Running (intermediate) value of the fraction value

keepGoing

boolean keepGoing

Flag to keep going or stop

Class gov.sandia.cognition.math.LogNumber extends Number implements Serializable

Serialized Fields

negative

boolean negative

The sign of the value, sign(value). True for negative, and false for positive.

logValue

double logValue

The log of the absolute value represented by this object, log(|value|).

Class gov.sandia.cognition.math.MutableDouble extends Number implements Serializable

serialVersionUID: 20110131L

Serialized Fields

value

double value

The value. Note: This is public just for performance reasons when people don't want to do the getter/setter for overhead reasons.

Class gov.sandia.cognition.math.MutableInteger extends Number implements Serializable

serialVersionUID: 20110601L

Serialized Fields

value

int value

The value. Note: This is public just for performance reasons when people don't want to do the getter/setter for overhead reasons.

Class gov.sandia.cognition.math.MutableLong extends Number implements Serializable

serialVersionUID: 20110602L

Serialized Fields

value

long value

The value. Note: This is public just for performance reasons when people don't want to do the getter/setter for overhead reasons.

Class gov.sandia.cognition.math.NumberAverager extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.OperationNotConvergedException extends RuntimeException implements Serializable

Class gov.sandia.cognition.math.RingAverager extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.UnivariateSummaryStatistics extends AbstractCloneableSerializable implements Serializable

Serialized Fields

min

double min

Min

max

double max

Max

quintiles

double[] quintiles

Quintile boundaries

confidenceLower

double confidenceLower

Lower 95% confidence region (alpha=0.025)

confidenceUpper

double confidenceUpper

Upper 95% confidence region (alpha=0.975)

median

double median

Median

numSamples

int numSamples

Number of samples

mean

double mean

Arithmetic mean

variance

double variance

Variance

skewness

double skewness

Skew

kurtosis

double kurtosis

Excess kurtosis

Class gov.sandia.cognition.math.UnsignedLogNumber extends Number implements Serializable

Serialized Fields

logValue

double logValue

The log of the value represented by this object, log(value).

Class gov.sandia.cognition.math.WeightedNumberAverager extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.WeightedRingAverager extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.math.geometry

Class gov.sandia.cognition.math.geometry.KDTree extends AbstractCollection<PairType extends Pair<? extends VectorType,DataType>> implements Serializable

Serialized Fields

num

int num

Number of elements in this subtree.

value

Pair<FirstType,SecondType> value

VectorType,DataType value for this node of the subtree.

parent

KDTree<VectorType extends Vectorizable,DataType,PairType extends Pair<? extends VectorType,DataType>> parent

Parent of this node of the subtree.

leftChild

KDTree<VectorType extends Vectorizable,DataType,PairType extends Pair<? extends VectorType,DataType>> leftChild

Left child of this subtree

rightChild

KDTree<VectorType extends Vectorizable,DataType,PairType extends Pair<? extends VectorType,DataType>> rightChild

Right child of this subtree.

comparator

KDTree.PairFirstVectorizableIndexComparator comparator

Comparator of this node to determine less than, greater than, or equality.

Class gov.sandia.cognition.math.geometry.KDTree.Neighborhood.Neighbor extends AbstractCloneableSerializable implements Serializable

Serialized Fields

pair

Pair<FirstType,SecondType> pair

Pair to store.

distance

double distance

Distance associated with this value.

Class gov.sandia.cognition.math.geometry.KDTree.PairFirstVectorizableIndexComparator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

comparator

VectorizableIndexComparator comparator

Embedded comparator for the Vectorizable argument.

Class gov.sandia.cognition.math.geometry.Quadtree extends AbstractCloneableSerializable implements Serializable

Serialized Fields

splitThreshold

int splitThreshold

The minimum number of items allowed in a leaf node. If there are more than this, then a node must be split. This number must be greater than zero.

items

LinkedList<E> items

All of the items in the tree. It should never be null.

initalBounds

Rectangle2D.Double initalBounds

The initial bounds for the tree. This may be null if they are not specified.

root

Quadtree.Node root

The root node of the tree. It should never be null.

Class gov.sandia.cognition.math.geometry.Quadtree.Node extends AbstractCloneableSerializable implements Serializable

Serialized Fields

parent

Quadtree.Node parent

The parent of this node in the tree. Null only for the root node.

bounds

Rectangle2D.Double bounds

The two-dimensional bounds for this node. This is only null if it is the root node and has no elements and no default bounds.

depth

int depth

The depth of this node in the tree.

localItems

LinkedList<E> localItems

The local items stored at this node.

lowerRight

Quadtree.Node lowerRight

The child for the lower-right quadrant of this node.

lowerLeft

Quadtree.Node lowerLeft

The child for the lower-left quadrant of this node.

upperLeft

Quadtree.Node upperLeft

The child for the upper-left quadrant of this node.

upperRight

Quadtree.Node upperRight

The child for the upper-right quadrant of this node.

children

ArrayList<E> children

The list of children for this node. Null to indicate that it has no children and thus is a leaf node.

Package gov.sandia.cognition.math.matrix

Class gov.sandia.cognition.math.matrix.AbstractMatrix extends AbstractRing<Matrix> implements Serializable

Class gov.sandia.cognition.math.matrix.AbstractVector extends AbstractVectorSpace<Vector,VectorEntry> implements Serializable

Class gov.sandia.cognition.math.matrix.AbstractVectorSpace extends AbstractRing<VectorType extends VectorSpace<VectorType,? extends EntryType>> implements Serializable

Class gov.sandia.cognition.math.matrix.DefaultInfiniteVector extends AbstractMutableDoubleMap<KeyType> implements Serializable

Class gov.sandia.cognition.math.matrix.DefaultVectorFactoryContainer extends AbstractCloneableSerializable implements Serializable

Serialized Fields

vectorFactory

VectorFactory<VectorType extends Vector> vectorFactory

The vector factory used to create new vectors.

Class gov.sandia.cognition.math.matrix.DimensionalityMismatchException extends RuntimeException implements Serializable

Class gov.sandia.cognition.math.matrix.MatrixFactory extends Object implements Serializable

Class gov.sandia.cognition.math.matrix.NumericalDifferentiator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

delta

double delta

Value for x-value differencing, must be greater than 0.0

internalFunction

Evaluator<InputType,OutputType> internalFunction

Internal function to numerically differencing.

Class gov.sandia.cognition.math.matrix.NumericalDifferentiator.DoubleJacobian extends NumericalDifferentiator<Double,Double,Double> implements Serializable

Class gov.sandia.cognition.math.matrix.NumericalDifferentiator.MatrixJacobian extends NumericalDifferentiator<Vector,Vector,Matrix> implements Serializable

Class gov.sandia.cognition.math.matrix.NumericalDifferentiator.VectorJacobian extends NumericalDifferentiator<Vector,Double,Vector> implements Serializable

Class gov.sandia.cognition.math.matrix.SparseVectorFactory extends VectorFactory<VectorType extends Vector> implements Serializable

Class gov.sandia.cognition.math.matrix.VectorFactory extends Object implements Serializable

Class gov.sandia.cognition.math.matrix.VectorizableIndexComparator extends AbstractCloneableSerializable implements Serializable

Serialized Fields

index

int index

Index to compare against.

Package gov.sandia.cognition.math.matrix.decomposition

Class gov.sandia.cognition.math.matrix.decomposition.EigenvectorPowerIteration extends AbstractCloneableSerializable implements Serializable

Package gov.sandia.cognition.math.matrix.mtj

Class gov.sandia.cognition.math.matrix.mtj.AbstractMTJMatrix extends AbstractMatrix implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.AbstractMTJVector extends AbstractVector implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.AbstractSparseMatrix extends AbstractMTJMatrix implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.DenseMatrix extends AbstractMTJMatrix implements Serializable

Serialization Methods

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

Reads in a serialized class from the specified stream

Throws:

IOException - On bad read

ClassNotFoundException - if next object isn't DenseMatrix

writeObject

private void writeObject(ObjectOutputStream out)
                  throws IOException

Writes a DenseMatrix out to a serialized file

Throws:

IOException - On bad write

Class gov.sandia.cognition.math.matrix.mtj.DenseMatrixFactoryMTJ extends MatrixFactory<DenseMatrix> implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.DenseVector extends AbstractMTJVector implements Serializable

Serialization Methods

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

Reads in a serialized class from the specified stream

Throws:

IOException - On bad read

ClassNotFoundException - if next object isn't DenseVector

writeObject

private void writeObject(ObjectOutputStream out)
                  throws IOException

Writes a DenseVector out to a serialized stream (usually file)

Throws:

IOException - On bad write

Class gov.sandia.cognition.math.matrix.mtj.DenseVectorFactoryMTJ extends VectorFactory<DenseVector> implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.DiagonalMatrixFactoryMTJ extends MatrixFactory<DiagonalMatrixMTJ> implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.DiagonalMatrixMTJ extends AbstractMTJMatrix implements Serializable

Serialization Methods

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

Reads in a serialized class from the specified stream

Throws:

IOException - On bad read

ClassNotFoundException - if next object isn't DenseMatrix

writeObject

private void writeObject(ObjectOutputStream out)
                  throws IOException

Writes a DenseMatrix out to a serialized file

Throws:

IOException - On bad write

Class gov.sandia.cognition.math.matrix.mtj.SparseColumnMatrix extends AbstractSparseMatrix implements Serializable

Serialization Methods

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

Custom deserialization is needed.

Throws:

IOException - If there is an error with the stream.

ClassNotFoundException - If a class used by this one cannot be found.

writeObject

private void writeObject(ObjectOutputStream out)
                  throws IOException

Custom serialization is needed.

Throws:

IOException - If there is an error writing to the stream.

Class gov.sandia.cognition.math.matrix.mtj.SparseMatrix extends SparseRowMatrix implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.SparseMatrixFactoryMTJ extends MatrixFactory<SparseMatrix> implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.SparseRowMatrix extends AbstractSparseMatrix implements Serializable

Serialization Methods

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

Custom deserialization is needed.

Throws:

IOException - If there is an error with the stream.

ClassNotFoundException - If a class used by this one cannot be found.

writeObject

private void writeObject(ObjectOutputStream out)
                  throws IOException

Custom serialization is needed.

Throws:

IOException - If there is an error writing to the stream.

Class gov.sandia.cognition.math.matrix.mtj.SparseVector extends AbstractMTJVector implements Serializable

Serialization Methods

readObject

private void readObject(ObjectInputStream in)
                 throws IOException,
                        ClassNotFoundException

This method provides custom deserialization for the class since the MTJ class does not implement Serializable.

Throws:

IOException - If there is an error reading the object.

ClassNotFoundException - If a class cannot be found.

writeObject

private void writeObject(ObjectOutputStream out)
                  throws IOException

This method provides custom serialization for the class since the MTJ class does not implement Serializable.

Throws:

IOException - If there is an error with the stream.

Class gov.sandia.cognition.math.matrix.mtj.SparseVectorFactoryMTJ extends SparseVectorFactory<SparseVector> implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.Vector1 extends DenseVector implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.Vector2 extends DenseVector implements Serializable

Class gov.sandia.cognition.math.matrix.mtj.Vector3 extends DenseVector implements Serializable

Package gov.sandia.cognition.math.matrix.mtj.decomposition

Class gov.sandia.cognition.math.matrix.mtj.decomposition.CholeskyDecompositionMTJ extends AbstractCloneableSerializable implements Serializable

Serialized Fields

R

DenseMatrix R

Cholesky factor, such that R.transpose().times( R ) equals the original matrix

Package gov.sandia.cognition.math.signals

Class gov.sandia.cognition.math.signals.AutoRegressiveMovingAverageFilter extends AbstractStatefulEvaluator<Double,Double,DefaultPair<FiniteCapacityBuffer<Double>,FiniteCapacityBuffer<Double>>> implements Serializable

Serialized Fields

movingAverageCoefficients

Vector movingAverageCoefficients

Coefficients of the moving-average filter. Element 0 is applied to the most-recent input, Element 1 is applied to the second-most-recent, and so forth. The dimensionality of the Vector is the order of the filter.

autoRegressiveCoefficients

Vector autoRegressiveCoefficients

Coefficients of the autoregressive filter. Element 0 is applied to the most-recent output, Element 1 is applied to the second-most-recent, and so forth. The dimensionality of the Vector is the order of the filter.

Class gov.sandia.cognition.math.signals.FourierTransform extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.signals.FourierTransform.Inverse extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.math.signals.LinearDynamicalSystem extends AbstractStatefulEvaluator<Vector,Vector,Vector> implements Serializable

Serialized Fields

A

Matrix A

System (Jacobian) matrix. Must be square.

B

Matrix B

Input-gain matrix. Columns must equal A's rows.

C

Matrix C

Output-selector matrix. Columns must equal A's rows.

Class gov.sandia.cognition.math.signals.MovingAverageFilter extends AbstractStatefulEvaluator<Double,Double,FiniteCapacityBuffer<Double>> implements Serializable

Serialized Fields

movingAverageCoefficients

Vector movingAverageCoefficients

Coefficients of the moving-average filter. Element 0 is applied to the most-recent input, Element 1 is applied to the second-most-recent, and so forth. The dimensionality of the Vector is the order of the filter.

Class gov.sandia.cognition.math.signals.PIDController extends AbstractStatefulEvaluator<Double,Double,PIDController.State> implements Serializable

Serialized Fields

targetInput

double targetInput

Set point target to achieve at steady-state.

proportionalGain

double proportionalGain

Proportional-error gain.

integralGain

double integralGain

Integral-error gain.

derivativeGain

double derivativeGain

Derivative-error gain.

Class gov.sandia.cognition.math.signals.PIDController.State extends AbstractCloneableSerializable implements Serializable

Serialized Fields

lastErr

double lastErr

Last error.

errSum

double errSum

Sum of all errors.

Package gov.sandia.cognition.statistics

Class gov.sandia.cognition.statistics.AbstractClosedFormSmoothUnivariateDistribution extends AbstractClosedFormUnivariateDistribution<Double> implements Serializable

Class gov.sandia.cognition.statistics.AbstractClosedFormUnivariateDistribution extends AbstractDistribution<NumberType extends Number> implements Serializable

Class gov.sandia.cognition.statistics.AbstractDataDistribution extends AbstractMutableDoubleMap<KeyType> implements Serializable

Class gov.sandia.cognition.statistics.AbstractDistribution extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.statistics.AbstractIncrementalEstimator extends AbstractBatchAndIncrementalLearner<DataType,SufficientStatisticsType extends SufficientStatistic<DataType,DistributionType>> implements Serializable

Class gov.sandia.cognition.statistics.AbstractRandomVariable extends AbstractRing<RandomVariable<DataType>> implements Serializable

Class gov.sandia.cognition.statistics.AbstractSufficientStatistic extends AbstractCloneableSerializable implements Serializable

Serialized Fields

count

long count

Number of data points used to create this SufficientStatistic

Class gov.sandia.cognition.statistics.DefaultDistributionParameter extends AbstractNamed implements Serializable

Serialized Fields

conditionalDistribution

ClosedFormDistribution<DataType> conditionalDistribution

Distribution from which to pull the parameters.

Class gov.sandia.cognition.statistics.UnivariateRandomVariable extends AbstractRandomVariable<Number> implements Serializable

Serialized Fields

numSamples

int numSamples

Number of samples to draw from the distribution to perform the empirical algebra approximation.

random

Random random

Random number generator used for sampling the distribution.

distribution

UnivariateDistribution<NumberType extends Number> distribution

Scalar distribution that backs the random variable, from which samples will be drawn to approximate the distribution during algebra.

Package gov.sandia.cognition.statistics.bayesian

Class gov.sandia.cognition.statistics.bayesian.AbstractBayesianParameter extends AbstractNamed implements Serializable

Serialized Fields

conditionalDistribution

ClosedFormDistribution<DataType> conditionalDistribution

Distribution from which to pull the parameters.

parameterPrior

Distribution<DataType> parameterPrior

Distribution of values that the parameter is assumed to take.

Class gov.sandia.cognition.statistics.bayesian.AbstractKalmanFilter extends AbstractBatchAndIncrementalLearner<Vector,MultivariateGaussian> implements Serializable

Serialized Fields

currentInput

Vector currentInput

Current input to the model.

modelCovariance

Matrix modelCovariance

Covariance associated with the system's model.

measurementCovariance

Matrix measurementCovariance

Covariance associated with the measurements.

Class gov.sandia.cognition.statistics.bayesian.AbstractMarkovChainMonteCarlo extends AbstractAnytimeBatchLearner<Collection<? extends ObservationType>,DataDistribution<ParameterType>> implements Serializable

Serialized Fields

random

Random random

Random number generator.

burnInIterations

int burnInIterations

The number of iterations that must transpire before the algorithm begins collection the samples.

iterationsPerSample

int iterationsPerSample

The number of iterations that must transpire between capturing samples from the distribution.

currentParameter

Object currentParameter

The current parameters in the random walk.

previousParameter

Object previousParameter

The previous parameter in the random walk.

Class gov.sandia.cognition.statistics.bayesian.AbstractParticleFilter extends AbstractBatchAndIncrementalLearner<ObservationType,DataDistribution<ParameterType>> implements Serializable

Serialized Fields

updater

ParticleFilter.Updater<ObservationType,ParameterType> updater

Updates the particle given an existing particle.

numParticles

int numParticles

Number of particles in the filter.

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling extends AbstractCloneableSerializable implements Serializable

Serialized Fields

logFunction

AdaptiveRejectionSampling.LogEvaluator<EvaluatorType extends Evaluator<Double,Double>> logFunction

Logarithm of the function that we want to evaluate

points

ArrayList<E> points

Input-output point pairs, sorted in ascending order by their x-axis value

maxNumPoints

int maxNumPoints

Maximum number of points that will be stored

minSupport

double minSupport

Minimum support (x-value) of the logFunction

maxSupport

double maxSupport

Maximum support (x-value) of the logFunction

upperEnvelope

AdaptiveRejectionSampling.UpperEnvelope upperEnvelope

Upper envelope of the logFunction

lowerEnvelope

AdaptiveRejectionSampling.LowerEnvelope lowerEnvelope

Lower envelope of the logFunction

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.AbstractEnvelope extends AbstractUnivariateScalarFunction implements Serializable

Serialized Fields

lines

ArrayList<E> lines

Line segments that comprise the envelope

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.LineSegment extends PolynomialFunction.Linear implements Serializable

Serialized Fields

left

double left

Left (minimum) x-axis value

right

double right

Right (maximum) x-axis value

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.LogEvaluator extends AbstractUnivariateScalarFunction implements Serializable

Serialized Fields

function

Evaluator<InputType,OutputType> function

Evaluator to wrap and compute the natural logarithm of.

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.LowerEnvelope extends AdaptiveRejectionSampling.AbstractEnvelope implements Serializable

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.PDFLogEvaluator extends AdaptiveRejectionSampling.LogEvaluator<ProbabilityFunction<Double>> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.Point extends DefaultInputOutputPair<Double,Double> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.AdaptiveRejectionSampling.UpperEnvelope extends AdaptiveRejectionSampling.AbstractEnvelope implements Serializable

Serialized Fields

segmentCDF

double[] segmentCDF

Cumulative sums of the normalized weights of the lines... This is automatically computed by computeSegments method.

Class gov.sandia.cognition.statistics.bayesian.BayesianCredibleInterval extends ConfidenceInterval implements Serializable

Class gov.sandia.cognition.statistics.bayesian.BayesianLinearRegression extends AbstractCloneableSerializable implements Serializable

Serialized Fields

outputVariance

double outputVariance

Assumed known variance of the outputs (measurements), must be greater than zero.

weightPrior

MultivariateGaussian weightPrior

Prior distribution of the weights, typically a zero-mean, diagonal-variance distribution.

Class gov.sandia.cognition.statistics.bayesian.BayesianLinearRegression.IncrementalEstimator extends BayesianLinearRegression implements Serializable

Class gov.sandia.cognition.statistics.bayesian.BayesianLinearRegression.IncrementalEstimator.SufficientStatistic extends AbstractSufficientStatistic<InputOutputPair<? extends Vectorizable,Double>,MultivariateGaussian> implements Serializable

Serialized Fields

z

Vector z

"z" statistic, proportional to the mean

covarianceInverse

Matrix covarianceInverse

Covariance inverse, sometimes called "precision"

Class gov.sandia.cognition.statistics.bayesian.BayesianLinearRegression.PredictiveDistribution extends AbstractCloneableSerializable implements Serializable

Serialized Fields

posterior

MultivariateGaussian posterior

Posterior distribution of the weights given the data.

Class gov.sandia.cognition.statistics.bayesian.BayesianRobustLinearRegression extends AbstractCloneableSerializable implements Serializable

Serialized Fields

weightPrior

MultivariateGaussian weightPrior

Prior distribution of the weights, typically a zero-mean, diagonal-variance distribution.

outputVariance

InverseGammaDistribution outputVariance

Distribution of the output (measurement) variance

Class gov.sandia.cognition.statistics.bayesian.BayesianRobustLinearRegression.IncrementalEstimator extends BayesianRobustLinearRegression implements Serializable

Class gov.sandia.cognition.statistics.bayesian.BayesianRobustLinearRegression.IncrementalEstimator.SufficientStatistic extends AbstractSufficientStatistic<InputOutputPair<? extends Vectorizable,Double>,MultivariateGaussianInverseGammaDistribution> implements Serializable

Serialized Fields

outputSumSquared

double outputSumSquared

Sum of the output squared

z

Vector z

"z" statistic, proportional to the mean

covarianceInverse

Matrix covarianceInverse

Covariance inverse, sometimes called "precision"

Class gov.sandia.cognition.statistics.bayesian.BayesianRobustLinearRegression.PredictiveDistribution extends AbstractCloneableSerializable implements Serializable

Serialized Fields

posterior

MultivariateGaussianInverseGammaDistribution posterior

Posterior distribution of the weights given the data.

Class gov.sandia.cognition.statistics.bayesian.DefaultBayesianParameter extends DefaultDistributionParameter<ParameterType,ConditionalType extends ClosedFormDistribution<?>> implements Serializable

Serialized Fields

parameterPrior

Distribution<DataType> parameterPrior

Distribution of values that the parameter is assumed to take.

Class gov.sandia.cognition.statistics.bayesian.DirichletProcessMixtureModel extends AbstractMarkovChainMonteCarlo<ObservationType,DirichletProcessMixtureModel.Sample<ObservationType>> implements Serializable

Serialized Fields

updater

DirichletProcessMixtureModel.Updater<ObservationType> updater

Creates the clusters and predictive prior distributions

numInitialClusters

int numInitialClusters

Number of clusters to initialize

reestimateAlpha

boolean reestimateAlpha

Flag to automatically re-estimate the alpha parameter

initialAlpha

double initialAlpha

Initial value of alpha, the concentration parameter of the Dirichlet Process

Class gov.sandia.cognition.statistics.bayesian.DirichletProcessMixtureModel.DPMMCluster extends DefaultCluster<ObservationType> implements Serializable

Serialized Fields

probabilityFunction

ProbabilityFunction<DataType> probabilityFunction

Probability function describing the assigned data

Class gov.sandia.cognition.statistics.bayesian.DirichletProcessMixtureModel.DPMMLogConditional extends AbstractCloneableSerializable implements Serializable

Serialized Fields

logConditional

double logConditional

log conditional likelihood

Class gov.sandia.cognition.statistics.bayesian.DirichletProcessMixtureModel.MultivariateMeanCovarianceUpdater extends AbstractCloneableSerializable implements Serializable

Serialized Fields

estimator

MultivariateGaussianMeanCovarianceBayesianEstimator estimator

Bayesian estimator for the parameters

Class gov.sandia.cognition.statistics.bayesian.DirichletProcessMixtureModel.MultivariateMeanUpdater extends AbstractCloneableSerializable implements Serializable

Serialized Fields

estimator

MultivariateGaussianMeanBayesianEstimator estimator

Bayesian estimator for the parameters

Class gov.sandia.cognition.statistics.bayesian.DirichletProcessMixtureModel.Sample extends AbstractCloneableSerializable implements Serializable

Serialized Fields

alpha

double alpha

Scaling parameter which defines the strength of the base distribution, must be greater than zero.

clusters

ArrayList<E> clusters

Point mass realizations from the base distribution.

posteriorLogLikelihood

Double posteriorLogLikelihood

Posterior log likelihood of the sample

Class gov.sandia.cognition.statistics.bayesian.ExtendedKalmanFilter extends AbstractKalmanFilter implements Serializable

Serialized Fields

motionModel

StatefulEvaluator<InputType,OutputType,StateType extends CloneableSerializable> motionModel

Model that determines how inputs and the previous state are updated.

observationModel

Evaluator<InputType,OutputType> observationModel

Model that determines how the state is observed.

Class gov.sandia.cognition.statistics.bayesian.GaussianProcessRegression extends DefaultKernelContainer<InputType> implements Serializable

Serialized Fields

outputVariance

double outputVariance

Assumed known variance of the outputs (measurements), must be greater than or equal to zero.

Class gov.sandia.cognition.statistics.bayesian.GaussianProcessRegression.PredictiveDistribution extends AbstractCloneableSerializable implements Serializable

Serialized Fields

inputs

ArrayList<E> inputs

Inputs that we've condition on.

posterior

MultivariateGaussian posterior

Posterior distribution of the Gaussian process given the data.

Class gov.sandia.cognition.statistics.bayesian.ImportanceSampling extends AbstractRandomized implements Serializable

Serialized Fields

updater

ImportanceSampling.Updater<ObservationType,ParameterType> updater

Updater for the ImportanceSampling algorithm.

numSamples

int numSamples

Number of samples.

Class gov.sandia.cognition.statistics.bayesian.ImportanceSampling.DefaultUpdater extends AbstractCloneableSerializable implements Serializable

Serialized Fields

conjuctive

BayesianParameter<ParameterType,ConditionalType extends Distribution<?>,PriorType extends Distribution<ParameterType>> conjuctive

Defines the parameter that connects the conditional and prior distributions.

Class gov.sandia.cognition.statistics.bayesian.KalmanFilter extends AbstractKalmanFilter implements Serializable

Serialized Fields

model

LinearDynamicalSystem model

Motion model of the underlying system.

Class gov.sandia.cognition.statistics.bayesian.MetropolisHastingsAlgorithm extends AbstractMarkovChainMonteCarlo<ObservationType,ParameterType> implements Serializable

Serialized Fields

currentLogLikelihood

double currentLogLikelihood

Log Likelihood of the current parameters.

updater

MetropolisHastingsAlgorithm.Updater<ObservationType,ParameterType> updater

The object that makes proposal samples from the current location.

Class gov.sandia.cognition.statistics.bayesian.ParallelDirichletProcessMixtureModel extends DirichletProcessMixtureModel<ObservationType> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.ParallelDirichletProcessMixtureModel.ClusterUpdaterTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

observations

Collection<E> observations

Observations that comprise the cluster

localUpdater

DirichletProcessMixtureModel.Updater<ObservationType> localUpdater

Local clone of the updater, needed to ensure thread safety

Class gov.sandia.cognition.statistics.bayesian.ParallelDirichletProcessMixtureModel.ObservationAssignmentTask extends AbstractCloneableSerializable implements Serializable

Serialized Fields

observations

Collection<E> observations

Observations to assign

weights

double[] weights

Weights that are re-used

assignments

ArrayList<E> assignments

Resulting assignments

logConditional

DirichletProcessMixtureModel.DPMMLogConditional logConditional

Log conditional of the previous sample

Class gov.sandia.cognition.statistics.bayesian.RejectionSampling extends AbstractRandomized implements Serializable

Serialized Fields

numSamples

int numSamples

Number of samples.

updater

RejectionSampling.Updater<ObservationType,ParameterType> updater

Updater for the ImportanceSampling algorithm.

Class gov.sandia.cognition.statistics.bayesian.RejectionSampling.DefaultUpdater extends AbstractCloneableSerializable implements Serializable

Serialized Fields

conjuctive

BayesianParameter<ParameterType,ConditionalType extends Distribution<?>,PriorType extends Distribution<ParameterType>> conjuctive

Defines the parameter that connects the conditional and prior distributions.

scale

Double scale

Scale factor to multiply the sampler function by to envelop the conjunctive distribution.

sampler

ProbabilityFunction<DataType> sampler

Distribution from which we sample and envelop the conjunctive distribution.

proposals

int proposals

Number of proposals suggested

Class gov.sandia.cognition.statistics.bayesian.SamplingImportanceResamplingParticleFilter extends AbstractParticleFilter<ObservationType,ParameterType> implements Serializable

Serialized Fields

particlePctThreadhold

double particlePctThreadhold

Percentage of effective particles, below which we resample.

Package gov.sandia.cognition.statistics.bayesian.conjugate

Class gov.sandia.cognition.statistics.bayesian.conjugate.AbstractConjugatePriorBayesianEstimator extends AbstractBatchAndIncrementalLearner<ObservationType,BeliefType extends ClosedFormDistribution<ParameterType>> implements Serializable

Serialized Fields

parameter

BayesianParameter<ParameterType,ConditionalType extends Distribution<?>,PriorType extends Distribution<ParameterType>> parameter

Bayesian hyperparameter relationship between the conditional distribution and the conjugate prior distribution.

Class gov.sandia.cognition.statistics.bayesian.conjugate.BernoulliBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Number,Double,BernoulliDistribution,BetaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.BernoulliBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,BernoulliDistribution,BetaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.BinomialBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Number,Double,BinomialDistribution,BetaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.BinomialBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,BinomialDistribution,BetaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.ExponentialBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Double,Double,ExponentialDistribution,GammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.ExponentialBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,ExponentialDistribution,GammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.GammaInverseScaleBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Double,Double,GammaDistribution,GammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.GammaInverseScaleBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,GammaDistribution,GammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.MultinomialBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Vector,Vector,MultinomialDistribution,DirichletDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.MultinomialBayesianEstimator.Parameter extends AbstractBayesianParameter<Vector,MultinomialDistribution,DirichletDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.MultivariateGaussianMeanBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Vector,Vector,MultivariateGaussian,MultivariateGaussian> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.MultivariateGaussianMeanBayesianEstimator.Parameter extends AbstractBayesianParameter<Vector,MultivariateGaussian,MultivariateGaussian> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.MultivariateGaussianMeanCovarianceBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Vector,Matrix,MultivariateGaussian,NormalInverseWishartDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.MultivariateGaussianMeanCovarianceBayesianEstimator.Parameter extends AbstractBayesianParameter<Matrix,MultivariateGaussian,NormalInverseWishartDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.PoissonBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Number,Double,PoissonDistribution,GammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.PoissonBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,PoissonDistribution,GammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.UniformDistributionBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Double,Double,UniformDistribution,ParetoDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.UniformDistributionBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,UniformDistribution,ParetoDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.UnivariateGaussianMeanBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Double,Double,UnivariateGaussian,UnivariateGaussian> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.UnivariateGaussianMeanBayesianEstimator.Parameter extends AbstractBayesianParameter<Double,UnivariateGaussian,UnivariateGaussian> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.UnivariateGaussianMeanVarianceBayesianEstimator extends AbstractConjugatePriorBayesianEstimator<Double,Vector,UnivariateGaussian,NormalInverseGammaDistribution> implements Serializable

Class gov.sandia.cognition.statistics.bayesian.conjugate.UnivariateGaussianMeanVarianceBayesianEstimator.Parameter extends AbstractBayesianParameter<Vector,UnivariateGaussian,NormalInverseGammaDistribution> implements Serializable

Package gov.sandia.cognition.statistics.distribution

Class gov.sandia.cognition.statistics.distribution.BernoulliDistribution extends AbstractClosedFormUnivariateDistribution<Number> implements Serializable

Serialized Fields

p

double p

Bernoulli parameter, where the distribution takes value "1" with probability "p" and value "0" with probability 1-p.

Class gov.sandia.cognition.statistics.distribution.BernoulliDistribution.CDF extends BernoulliDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BernoulliDistribution.PMF extends BernoulliDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaBinomialDistribution extends AbstractClosedFormUnivariateDistribution<Number> implements Serializable

Serialized Fields

n

int n

Number of observations, similar to the Binomial N, must be greater than zero

shape

double shape

Shape, similar to the beta parameter shape, must be greater than zero

scale

double scale

Scale, similar to the beta parameter scale, must be greater than zero

Class gov.sandia.cognition.statistics.distribution.BetaBinomialDistribution.CDF extends BetaBinomialDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaBinomialDistribution.MomentMatchingEstimator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaBinomialDistribution.PMF extends BetaBinomialDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaDistribution extends AbstractClosedFormSmoothUnivariateDistribution implements Serializable

Serialized Fields

alpha

double alpha

Alpha shape parameter, must be greater than 0 (typically greater than 1)

beta

double beta

Alpha shape parameter, must be greater than 0 (typically greater than 1)

Class gov.sandia.cognition.statistics.distribution.BetaDistribution.CDF extends BetaDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaDistribution.MomentMatchingEstimator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaDistribution.PDF extends BetaDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BetaDistribution.WeightedMomentMatchingEstimator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.statistics.distribution.BinomialDistribution extends AbstractClosedFormUnivariateDistribution<Number> implements Serializable

Serialized Fields

N

int N

Total number of experiments, must be greater than zero

p

double p

Probability of a positive outcome (Bernoulli probability), [0,1]

Class gov.sandia.cognition.statistics.distribution.BinomialDistribution.CDF extends BinomialDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.BinomialDistribution.MaximumLikelihoodEstimator extends AbstractCloneableSerializable implements Serializable

Class gov.sandia.cognition.statistics.distribution.BinomialDistribution.PMF extends BinomialDistribution implements Serializable

Class gov.sandia.cognition.statistics.distribution.CategoricalDistribution extends AbstractDistribution<Vector> implements Serializable

Serialized Fields

parameters

Vector parameters

Par