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

Provides ensmble methods.

See:
          Description

Interface Summary
Ensemble<MemberType> The Ensemble interface defines the functionality of an "ensemble" that is typically created by combining together the result of multiple learning algorithms.
 

Class Summary
AbstractBaggingLearner<InputType,OutputType,MemberType,EnsembleType extends Evaluator<? super InputType,? extends OutputType>> Learns an ensemble by randomly sampling with replacement (duplicates allowed) some percentage of the size of the data (defaults to 100%) on each iteration to train a new ensemble member.
AbstractUnweightedEnsemble<MemberType> An abstract implementation of the Ensemble interface for unweighted ensembles.
AbstractWeightedEnsemble<MemberType> An abstract implementation of the Ensemble interface for ensembles that have a weight associated with each member.
AdaBoost<InputType> The AdaBoost class implements the Adaptive Boosting (AdaBoost) algorithm formulated by Yoav Freund and Robert Shapire.
AdditiveEnsemble<InputType,MemberType extends Evaluator<? super InputType,? extends Number>> An ensemble of regression functions that determine the result by adding their outputs together.
AveragingEnsemble<InputType,MemberType extends Evaluator<? super InputType,? extends Number>> An ensemble for regression functions that averages together the output value of each ensemble member to get the final output.
BaggingCategorizerLearner<InputType,CategoryType> Learns an categorization ensemble by randomly sampling with replacement (duplicates allowed) some percentage of the size of the data (defaults to 100%) on each iteration to train a new ensemble member.
BaggingRegressionLearner<InputType> Learns an ensemble for regression by randomly sampling with replacement (duplicates allowed) some percentage of the size of the data (defaults to 100%) on each iteration to train a new ensemble member.
BinaryBaggingLearner<InputType> The BinaryBaggingLearner implements the Bagging learning algorithm.
BinaryCategorizerSelector<InputType> The BinaryCategorizerSelector class implements a "weak learner" meant for use in boosting algorithms that selects the best BinaryCategorizer from a pre-set list by picking the one with the best weighted error.
CategoryBalancedBaggingLearner<InputType,CategoryType> An extension of the basic bagging learner that attempts to sample bags that have equal numbers of examples from every category.
CategoryBalancedIVotingLearner<InputType,CategoryType> An extension of IVoting for dealing with skew problems that makes sure that there are an equal number of examples from each category in each sample that an ensemble member is trained on.
IVotingCategorizerLearner<InputType,CategoryType> Learns an ensemble in a method similar to bagging except that on each iteration the bag is built from two parts, each sampled from elements from disjoint sets.
IVotingCategorizerLearner.OutOfBagErrorStoppingCriteria<InputType,CategoryType> Implements a stopping criteria for IVoting that uses the out-of-bag error to determine when to stop learning the ensemble.
MultiCategoryAdaBoost<InputType,CategoryType> An implementation of a multi-class version of the Adaptive Boosting (AdaBoost) algorithm, known as AdaBoost.M1.
OnlineBaggingCategorizerLearner<InputType,CategoryType,MemberType extends Evaluator<? super InputType,? extends CategoryType>> An implementation of an online version of the Bagging algorithm for learning an ensemble of categorizers.
VotingCategorizerEnsemble<InputType,CategoryType,MemberType extends Evaluator<? super InputType,? extends CategoryType>> An ensemble of categorizers that determine the result based on an equal-weight vote.
WeightedAdditiveEnsemble<InputType,MemberType extends Evaluator<? super InputType,? extends Number>> An implementation of an ensemble that takes a weighted sum of the values returned by its members.
WeightedAveragingEnsemble<InputType,MemberType extends Evaluator<? super InputType,? extends Number>> An implementation of an ensemble that takes the weighted average of its members.
WeightedBinaryEnsemble<InputType,MemberType extends Evaluator<? super InputType,? extends Boolean>> The WeightedBinaryEnsemble class implements an Ensemble of BinaryCategorizer objects where each categorizer is assigned a weight and the category is selected by choosing the one with the largest sum of weights.
WeightedVotingCategorizerEnsemble<InputType,CategoryType,MemberType extends Evaluator<? super InputType,? extends CategoryType>> An ensemble of categorizers where each ensemble member is evaluated with the given input to find the category to which its weighted votes are assigned.
 

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

Provides ensmble methods. Ensemble methods is a general term used for a type of meta-learner that combines the results of one or more learning algorithms together to create an "ensemble" of learned objects.

Since:
2.0
Author:
Justin Basilico