Package org.neuroph.nnet.learning

Source Code of org.neuroph.nnet.learning.BinaryDeltaRule

/**
* Copyright 2010 Neuroph Project http://neuroph.sourceforge.net
*
* 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.
*/


package org.neuroph.nnet.learning;

import org.neuroph.core.NeuralNetwork;
import org.neuroph.core.Neuron;
import org.neuroph.nnet.comp.ThresholdNeuron;

/**
* Delta rule learning algorithm for perceptrons with step functions.
*
* The difference to Perceptronlearning is that Delta Rule calculates error
* before the non-lnear step transfer function
*
* @author Zoran Sevarac <sevarac@gmail.com>
*/
public class BinaryDeltaRule extends PerceptronLearning {

  /**
   * The class fingerprint that is set to indicate serialization
   * compatibility with a previous version of the class.
   */
  private static final long serialVersionUID = 1L;

  /**
   * The errorCorrection parametar of this learning algorithm
   */
  private double errorCorrection = 0.1;

  /**
   * Creates new BinaryDeltaRule learning
   */
  public BinaryDeltaRule() {
    super();
  }


  /**
   * This method implements weight update procedure for the whole network for
   * this learning rule
   *
   * @param patternError
   *            single pattern error vector
         *
         * if the output is 0 and required value is 1, increase rthe weights
         * if the output is 1 and required value is 0, decrease the weights
         * otherwice leave weights unchanged
         *
   */
  @Override
  protected void updateNetworkWeights(double[] patternError) {
    int i = 0;
    for(Neuron outputNeuron : neuralNetwork.getOutputNeurons()) {
      ThresholdNeuron neuron = (ThresholdNeuron)outputNeuron;
      double outputError = patternError[i];
      double thresh = neuron.getThresh();
      double netInput = neuron.getNetInput();
      double threshError =  thresh - netInput; // distance from zero
                        // use output error to decide weathet to inrease, decrase or leave unchanged weights
                        // add errorCorrection to threshError to move above or below zero
                        double neuronError = outputError * (Math.abs(threshError) + errorCorrection);

                        // use same adjustment principle as PerceptronLearning,
                        // just with different neuronError
                        neuron.setError(neuronError);
      updateNeuronWeights(neuron);

      i++;
    } // for
  }

        /**
   * Gets the errorCorrection parametar
   *
   * @return errorCorrection parametar
   */
  public double getErrorCorrection() {
    return this.errorCorrection;
  }

  /**
   * Sets the errorCorrection parametar
   *
   * @param errorCorrection
   *            the value for errorCorrection parametar
   */
  public void setErrorCorrection(double errorCorrection) {
    this.errorCorrection = errorCorrection;
  }

}
TOP

Related Classes of org.neuroph.nnet.learning.BinaryDeltaRule

TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.