de.jungblut.classification.regression

Class LogisticRegressionCostFunction

java.lang.Object

de.jungblut.classification.regression.LogisticRegressionCostFunction

All Implemented Interfaces:

CostFunction

public final class LogisticRegressionCostFunction
extends Object
implements CostFunction

Constructor Summary

Constructors

Constructor and Description
LogisticRegressionCostFunction(de.jungblut.math.DoubleMatrix x, de.jungblut.math.DoubleMatrix y, double lambda)

Method Summary

Modifier and Type	Method and Description
CostGradientTuple	evaluateCost(de.jungblut.math.DoubleVector theta) Evaluation for the cost function to retrieve cost and gradient.

Methods inherited from class java.lang.Object

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

Constructor Detail

LogisticRegressionCostFunction

public LogisticRegressionCostFunction(de.jungblut.math.DoubleMatrix x,
                                      de.jungblut.math.DoubleMatrix y,
                                      double lambda)

Parameters:

x - normal feature matrix, column 0 should contain the bias.

y - normal outcome matrix, for multiple classes use the one-hot encoding. This matrix should be transposed.

lambda - l1 reg parameter.

Method Detail

evaluateCost

public CostGradientTuple evaluateCost(de.jungblut.math.DoubleVector theta)

Description copied from interface: CostFunction

Evaluation for the cost function to retrieve cost and gradient.

Specified by:

evaluateCost in interface CostFunction

Parameters:

theta - a given input vector

Returns:

a tuple consists of J (cost) and a vector X which is the gradient of the input.