Package org.encog.ml.train

Source Code of org.encog.ml.train.BasicTraining

/*
* Encog(tm) Core v3.3 - Java Version
* http://www.heatonresearch.com/encog/
* https://github.com/encog/encog-java-core
* Copyright 2008-2014 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.ml.train;

import java.util.ArrayList;
import java.util.List;

import org.encog.ml.TrainingImplementationType;
import org.encog.ml.data.MLDataSet;
import org.encog.ml.train.strategy.Strategy;
import org.encog.ml.train.strategy.end.EndTrainingStrategy;

/**
* An abstract class that implements basic training for most training
* algorithms. Specifically training strategies can be added to enhance the
* training.
*
* @author jheaton
*
*/
public abstract class BasicTraining implements MLTrain {

  /**
   * The training strategies to use.
   */
  private final List<Strategy> strategies = new ArrayList<Strategy>();

  /**
   * The training data.
   */
  private MLDataSet training;

  /**
   * The current error rate.
   */
  private double error;

  /**
   * The current iteration.
   */
  private int iteration;
 
  private TrainingImplementationType implementationType;
 
  /**
   * Used for serialization.
   */
  public BasicTraining() {
   
  }

  public BasicTraining(TrainingImplementationType implementationType) {
    this.implementationType = implementationType;
  }
 
  /**
   * Training strategies can be added to improve the training results. There
   * are a number to choose from, and several can be used at once.
   *
   * @param strategy
   *            The strategy to add.
   */
  public void addStrategy(final Strategy strategy) {
    strategy.init(this);
    this.strategies.add(strategy);
  }

  /**
   * Should be called after training has completed and the iteration method
   * will not be called any further.
   */
  public void finishTraining() {

     
  }

  /**
   * {@inheritDoc}
   */
  public double getError() {
    return this.error;
  }

  /**
   * @return the iteration
   */
  public int getIteration() {
    return this.iteration;
  }

  /**
   * @return The strategies to use.
   */
  public List<Strategy> getStrategies() {
    return this.strategies;
  }

  /**
   * @return The training data to use.
   */
  public MLDataSet getTraining() {
    return this.training;
  }

  /**
   * @return True if training can progress no further.
   */
  public boolean isTrainingDone() {
    for (Strategy strategy : this.strategies) {
      if (strategy instanceof EndTrainingStrategy) {
        EndTrainingStrategy end = (EndTrainingStrategy)strategy;
        if( end.shouldStop() ) {
          return true;
        }
      }
    }
   
    return false;
  }

  /**
   * Perform the specified number of training iterations. This is a basic
   * implementation that just calls iteration the specified number of times.
   * However, some training methods, particularly with the GPU, benefit
   * greatly by calling with higher numbers than 1.
   *
   * @param count
   *            The number of training iterations.
   */
  public void iteration(final int count) {
    for (int i = 0; i < count; i++) {
      iteration();
    }
  }

  /**
   * Call the strategies after an iteration.
   */
  public void postIteration() {
    for (final Strategy strategy : this.strategies) {
      strategy.postIteration();
    }
  }

  /**
   * Call the strategies before an iteration.
   */
  public void preIteration() {

    this.iteration++;

    for (final Strategy strategy : this.strategies) {
      strategy.preIteration();
    }
  }

  /**
   * @param error
   *            Set the current error rate. This is usually used by training
   *            strategies.
   */
  public void setError(final double error) {
    this.error = error;
  }

  /**
   * @param iteration
   *            the iteration to set
   */
  public void setIteration(final int iteration) {
    this.iteration = iteration;
  }

  /**
   * Set the training object that this strategy is working with.
   *
   * @param training
   *            The training object.
   */
  public void setTraining(final MLDataSet training) {
    this.training = training;
  }
 
  public TrainingImplementationType getImplementationType() {
    return this.implementationType;
  }

}
TOP

Related Classes of org.encog.ml.train.BasicTraining

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.