gov.sandia.cognition.learning.function.cost

Class SumSquaredErrorCostFunction.Cache

java.lang.Object

gov.sandia.cognition.util.AbstractCloneableSerializable

gov.sandia.cognition.learning.function.cost.SumSquaredErrorCostFunction.Cache

All Implemented Interfaces:

CloneableSerializable, java.io.Serializable, java.lang.Cloneable

Enclosing class:

SumSquaredErrorCostFunction

public static class SumSquaredErrorCostFunction.Cache
extends AbstractCloneableSerializable

Caches often-used values for the Cost Function

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
Matrix	J Jacobian
Vector	Jte Jacobian transpose times Error: J.transpose().times( error )
Matrix	JtJ Inner-product of the Jacobian matrix: J.transpose().times( J )
double	parameterCost Cost-function value of the parameter set

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	Cache(Matrix J, Matrix JtJ, Vector Jte, double parameterCost) Creates a new instance of Cache

Method Summary

Modifier and Type	Method and Description
static SumSquaredErrorCostFunction.Cache	compute(GradientDescendable objectToOptimize, java.util.Collection<? extends InputOutputPair<? extends Vector,Vector>> data) Computes often-used parameters of a sum-squared error term

Methods inherited from class gov.sandia.cognition.util.AbstractCloneableSerializable

clone

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

J

public final Matrix J

Jacobian

JtJ

public final Matrix JtJ

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

Jte

public final Vector Jte

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

parameterCost

public final double parameterCost

Cost-function value of the parameter set

Constructor Detail

Cache

protected Cache(Matrix J,
                Matrix JtJ,
                Vector Jte,
                double parameterCost)

Creates a new instance of Cache

Parameters:

J - Jacobian

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

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

parameterCost - Cost-function value of the parameter set

Method Detail

compute

public static SumSquaredErrorCostFunction.Cache compute(GradientDescendable objectToOptimize,
                                                        java.util.Collection<? extends InputOutputPair<? extends Vector,Vector>> data)

Computes often-used parameters of a sum-squared error term

Parameters:

objectToOptimize - GradientDescendable to compute the statistics of

data - Dataset to consider

Returns:

Cache containing the cached cost-function parameters