SVDTraining.java

/*
 * Encog(tm) Core v3.4 - Java Version
 * http://www.heatonresearch.com/encog/
 * https://github.com/encog/encog-java-core
 
 * Copyright 2008-2016 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.neural.rbf.training;

import org.encog.mathutil.rbf.RadialBasisFunction;
import org.encog.ml.TrainingImplementationType;
import org.encog.ml.data.MLDataSet;
import org.encog.ml.train.BasicTraining;
import org.encog.neural.networks.training.TrainingError;
import org.encog.neural.networks.training.propagation.TrainingContinuation;
import org.encog.neural.rbf.RBFNetwork;
import org.encog.util.ObjectPair;
import org.encog.util.simple.TrainingSetUtil;

/**
 * Train a RBF neural network using a SVD.
 * 
 * Contributed to Encog By M.Fletcher and M.Dean University of Cambridge, Dept.
 * of Physics, UK
 * 
 */
public class SVDTraining extends BasicTraining {

	/**
	 * The network that is to be trained.
	 */
	private final RBFNetwork network;

	/**
	 * Construct the training object.
	 * 
	 * @param network
	 *            The network to train. Must have a single output neuron.
	 * @param training
	 *            The training data to use. Must be indexable.
	 */
	public SVDTraining(final RBFNetwork network, final MLDataSet training) {
		super(TrainingImplementationType.OnePass);
		if (network.getOutputCount() != 1) {
			throw new TrainingError(
					"SVD requires an output layer with a single neuron.");
		}

		setTraining(training);
		this.network = network;
	}

	@Override
	public boolean canContinue() {
		return false;
	}

	public void flatToMatrix(final double[] flat, final int start,
			final double[][] matrix) {
		final int rows = matrix.length;
		final int cols = matrix[0].length;

		int index = start;

		for (int r = 0; r < rows; r++) {
			for (int c = 0; c < cols; c++) {
				matrix[r][c] = flat[index++];
			}
		}
	}

	/**
	 * {@inheritDoc}
	 */
	@Override
	public RBFNetwork getMethod() {
		return this.network;
	}

	/**
	 * Perform one iteration.
	 */
	@Override
	public void iteration() {
		final int length = this.network.getRBF().length;

		final RadialBasisFunction[] funcs = new RadialBasisFunction[length];

		// Iteration over neurons and determine the necessaries
		for (int i = 0; i < length; i++) {
			final RadialBasisFunction basisFunc = this.network.getRBF()[i];

			funcs[i] = basisFunc;

			// This is the value that is changed using other training methods.
			// weights[i] =
			// network.Structure.Synapses[0].WeightMatrix.Data[i][j];
		}

		final ObjectPair<double[][], double[][]> data = TrainingSetUtil
				.trainingToArray(getTraining());

		final double[][] matrix = new double[length][this.network
				.getOutputCount()];

		flatToMatrix(this.network.getFlat().getWeights(), 0, matrix);
		setError(SVD.svdfit(data.getA(), data.getB(), matrix, funcs));
		matrixToFlat(matrix, this.network.getFlat().getWeights(), 0);
	}

	/**
	 * Convert the matrix to flat.
	 * @param matrix The matrix.
	 * @param flat Flat array.
	 * @param start WHere to start.
	 */
	public void matrixToFlat(final double[][] matrix, final double[] flat,
			final int start) {
		final int rows = matrix.length;
		final int cols = matrix[0].length;

		int index = start;

		for (int r = 0; r < rows; r++) {
			for (int c = 0; c < cols; c++) {
				flat[index++] = matrix[r][c];
			}
		}
	}

	/**
	 * {@inheritDoc}
	 */
	@Override
	public TrainingContinuation pause() {
		return null;
	}

	/**
	 * {@inheritDoc}
	 */
	@Override
	public void resume(final TrainingContinuation state) {

	}

}