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

Provides line (scalar) minimization algorithms.

See:
          Description

Interface Summary
LineMinimizer<EvaluatorType extends Evaluator<Double,Double>> Defines the functionality of a line-minimization algorithm, often called a "line search" algorithm.
 

Class Summary
AbstractAnytimeLineMinimizer<EvaluatorType extends Evaluator<Double,Double>> Partial AnytimeAlgorithm implementation of a LineMinimizer.
DirectionalVectorToDifferentiableScalarFunction Creates a truly differentiable scalar function from a differentiable Vector function, instead of using a forward-differences approximation to the derivative like DirectionalVectorToScalarFunction does.
DirectionalVectorToScalarFunction Maps a vector function onto a scalar one by using a directional vector and vector offset, and the parameter to the function is a scalar value along the direction from the start-point offset.
InputOutputSlopeTriplet Stores an InputOutputPair with corresponding slope (gradient) information
LineBracket Class that defines a bracket for a scalar function.
LineMinimizerBacktracking Implementation of the backtracking line-minimization algorithm.
LineMinimizerDerivativeBased This is an implementation of a line-minimization algorithm proposed by Fletcher that makes extensive use of first-order derivative information.
LineMinimizerDerivativeFree This is an implementation of a LineMinimizer that does not require derivative information.
WolfeConditions The Wolfe conditions define a set of sufficient conditions for "sufficient decrease" in inexact line search.
 

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

Provides line (scalar) minimization algorithms. These algorithms are primarily used as a subroutine in multivariate minimization algorithms.

Since:
2.1
Author:
Kevin R. Dixon