/*
* Encog(tm) Core v3.0 - Java Version
* http://www.heatonresearch.com/encog/
* http://code.google.com/p/encog-java/
* Copyright 2008-2011 Heaton Research, Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*
* For more information on Heaton Research copyrights, licenses
* and trademarks visit:
* http://www.heatonresearch.com/copyright
*/
package org.encog.engine.network.activation;
import org.encog.EncogError;
/**
* An activation function that only allows a specified number, usually one, of
* the out-bound connection to win. These connections will share in the sum of
* the output, whereas the other neurons will receive zero.
*
* This activation function can be useful for "winner take all" layers.
*
*/
public class ActivationCompetitive implements ActivationFunction {
/**
* The offset to the parameter that holds the max winners.
*/
public static final int PARAM_COMPETITIVE_MAX_WINNERS = 0;
/**
* The serial ID.
*/
private static final long serialVersionUID = 5396927873082336888L;
/**
* The parameters.
*/
private final double[] params;
/**
* Create a competitive activation function with one winner allowed.
*/
public ActivationCompetitive() {
this(1);
}
/**
* Create a competitive activation function with the specified maximum
* number of winners.
*
* @param winners
* The maximum number of winners that this function supports.
*/
public ActivationCompetitive(final int winners) {
this.params = new double[1];
this.params[
ActivationCompetitive.PARAM_COMPETITIVE_MAX_WINNERS] = winners;
}
/**
* {@inheritDoc}
*/
@Override
public final void activationFunction(final double[] x, final int start,
final int size) {
final boolean[] winners = new boolean[x.length];
double sumWinners = 0;
// find the desired number of winners
for (int i = 0; i < this.params[0]; i++) {
double maxFound = Double.NEGATIVE_INFINITY;
int winner = -1;
// find one winner
for (int j = start; j < start + size; j++) {
if (!winners[j] && (x[j] > maxFound)) {
winner = j;
maxFound = x[j];
}
}
sumWinners += maxFound;
winners[winner] = true;
}
// adjust weights for winners and non-winners
for (int i = start; i < start + size; i++) {
if (winners[i]) {
x[i] = x[i] / sumWinners;
} else {
x[i] = 0.0;
}
}
}
/**
* @return A cloned copy of this object.
*/
@Override
public final ActivationFunction clone() {
return new ActivationCompetitive(
(int) this.params[
ActivationCompetitive.PARAM_COMPETITIVE_MAX_WINNERS]);
}
/**
* Implements the activation function. The array is modified according to
* the activation function being used. See the class description for more
* specific information on this type of activation function.
*
* @param d
* The input array to the activation function.
* @return The derivative.
*/
@Override
public final double derivativeFunction(final double b, final double a) {
throw new EncogError("Can't use the competitive activation function "
+ "where a derivative is required.");
}
/**
* @return The maximum number of winners this function supports.
*/
public final int getMaxWinners() {
return (int) this.params[
ActivationCompetitive.PARAM_COMPETITIVE_MAX_WINNERS];
}
/**
* {@inheritDoc}
*/
@Override
public final String[] getParamNames() {
final String[] result = { "maxWinners" };
return result;
}
/**
* {@inheritDoc}
*/
@Override
public final double[] getParams() {
return this.params;
}
/**
* @return False, indication that no derivative is available for this
* function.
*/
@Override
public final boolean hasDerivative() {
return false;
}
/**
* {@inheritDoc}
*/
@Override
public final void setParam(final int index, final double value) {
this.params[index] = value;
}
}