/*
* RapidMiner
*
* Copyright (C) 2001-2011 by Rapid-I and the contributors
*
* Complete list of developers available at our web site:
*
* http://rapid-i.com
*
* This program is free software: you can redistribute it and/or modify
* it under the terms of the GNU Affero General Public License as published by
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU Affero General Public License for more details.
*
* You should have received a copy of the GNU Affero General Public License
* along with this program. If not, see http://www.gnu.org/licenses/.
*/
package com.rapidminer.operator.performance;
import java.util.Iterator;
import com.rapidminer.example.Attribute;
import com.rapidminer.example.Example;
import com.rapidminer.example.ExampleSet;
import com.rapidminer.operator.OperatorException;
import com.rapidminer.tools.math.Averagable;
import com.rapidminer.tools.math.MathFunctions;
/**
* Calculates the cross-entropy for the predictions of a classifier.
*
* @author Ingo Mierswa
*/
public class CrossEntropy extends MeasuredPerformance {
private static final long serialVersionUID = 8341971882780129465L;
/** The value of the criterion. */
private double value = Double.NaN;
private double counter = 1.0d;
/** Clone constructor. */
public CrossEntropy() {}
public CrossEntropy(CrossEntropy c) {
super(c);
this.value = c.value;
this.counter = c.counter;
}
/** Calculates the margin. */
@Override
public void startCounting(ExampleSet exampleSet, boolean useExampleWeights) throws OperatorException {
super.startCounting(exampleSet, useExampleWeights);
// compute margin
Iterator<Example> reader = exampleSet.iterator();
this.value = 0.0d;
Attribute labelAttr = exampleSet.getAttributes().getLabel();
Attribute weightAttribute = null;
if (useExampleWeights)
weightAttribute = exampleSet.getAttributes().getWeight();
while (reader.hasNext()) {
Example example = reader.next();
String trueLabel = example.getNominalValue(labelAttr);
double confidence = example.getConfidence(trueLabel);
double weight = 1.0d;
if (weightAttribute != null)
weight = example.getValue(weightAttribute);
this.value -= weight * MathFunctions.ld(confidence);
this.counter += weight;
}
}
/** Does nothing. Everything is done in {@link #startCounting(ExampleSet, boolean)}. */
@Override
public void countExample(Example example) {}
@Override
public double getExampleCount() {
return counter;
}
@Override
public double getMikroVariance() {
return Double.NaN;
}
@Override
public double getMikroAverage() {
return value / counter;
}
/** Returns the fitness. */
@Override
public double getFitness() {
return -1 * getAverage();
}
@Override
public String getName() {
return "cross-entropy";
}
@Override
public String getDescription() {
return "The cross-entropy of a classifier, defined as the sum over the logarithms of the true label's confidences divided by the number of examples";
}
@Override
public void buildSingleAverage(Averagable performance) {
CrossEntropy other = (CrossEntropy) performance;
this.value += other.value;
this.counter += other.counter;
}
/** Returns the super class implementation of toString(). */
@Override
public String toString() {
return super.toString();
}
}