org.openimaj.ml.linear.learner.matlib.loss

Class LossFunction

java.lang.Object

org.openimaj.ml.linear.learner.matlib.loss.LossFunction

Direct Known Subclasses:

MatLossFunction, MatSquareLossFunction

public abstract class LossFunction
extends Object

With a held Y and X, return gradient and evaluations of a loss function of some parameters s.t. J(W) = F(Y,X,W)

Author:

Sina Samangooei (ss@ecs.soton.ac.uk)

Field Summary

Fields

Modifier and Type	Field and Description
protected ch.akuhn.matrix.Matrix	bias
protected ch.akuhn.matrix.Matrix	X
protected ch.akuhn.matrix.Matrix	Y

Constructor Summary

Constructors

Constructor and Description
LossFunction()

Method Summary

Modifier and Type	Method and Description
abstract double	eval(ch.akuhn.matrix.Matrix W)
abstract ch.akuhn.matrix.Matrix	gradient(ch.akuhn.matrix.Matrix W)
abstract boolean	isMatrixLoss()
void	setBias(ch.akuhn.matrix.Matrix bias)
void	setX(ch.akuhn.matrix.Matrix X)
void	setY(ch.akuhn.matrix.Matrix Y)

Methods inherited from class java.lang.Object

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

Field Detail

X

protected ch.akuhn.matrix.Matrix X

Y

protected ch.akuhn.matrix.Matrix Y

bias

protected ch.akuhn.matrix.Matrix bias

Constructor Detail

LossFunction

public LossFunction()

Method Detail

setX

public void setX(ch.akuhn.matrix.Matrix X)

setY

public void setY(ch.akuhn.matrix.Matrix Y)

gradient

public abstract ch.akuhn.matrix.Matrix gradient(ch.akuhn.matrix.Matrix W)

eval

public abstract double eval(ch.akuhn.matrix.Matrix W)

setBias

public void setBias(ch.akuhn.matrix.Matrix bias)

isMatrixLoss

public abstract boolean isMatrixLoss()