linear.java

来自「利用Java实现的神经网络工具箱」· Java 代码 · 共 197 行

/*
 * $RCSfile: Linear.java,v $
 * $Revision: 1.8 $
 * $Date: 2004/10/24 03:05:29 $
 *
 * NeuralNetworkToolkit
 * Copyright (C) 2004 Universidade de Brasília
 *
 * This file is part of NeuralNetworkToolkit.
 *
 * NeuralNetworkToolkit is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 2 of the License, or
 * (at your option) any later version.
 *
 * NeuralNetworkToolkit 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 General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with NeuralNetworkToolkit; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA - 02111-1307 - USA.
 */

package neuralnetworktoolkit.normalization;

/**
 * Linear normalization (or "Normalization into an Interval").<br>
 * <br>
 * Implementation according the description avaiable in
 * <a href="http://www.ivorix.com/en/products/tech/norm/">
 * http://www.ivorix.com/en/products/tech/norm/</a>.
 * 
 * @version $Revision: 1.8 $ - $Date: 2004/10/24 03:05:29 $
 * 
 * @author <a href="mailto:[e-mail]">Rodrigo C. M. Coimbra</a>
 * @author <a href="mailto:[e-mail]">Hugo Iver V. Gon�alves</a>
 */
public class Linear extends Normalization {
	private double[] upperLimit;
	private double[] lowerLimit;
	private double[] upperValue;
	private double[] lowerValue;

	/* 
	 * @see neuralnetworktoolkit.normalization.INormalization#normalize(double[][])
	 */
	public double[][] normalize(double[][] originalData) {
		double[][] normalizedData;
		
		if ((upperLimit == null) || (lowerLimit == null)
				|| (upperValue == null) || (lowerValue == null)) {

			setupParameters(originalData);
			
		}
		normalizedData = transform(originalData);
		
		return normalizedData;
		
	} //normalize()
	
	/* 
	 * @see neuralnetworktoolkit.normalization.INormalization#unnormalize(double[][])
	 */
	public double[][] unnormalize(double[][] normalizedData) {
		double[][] originalData;
		
		originalData = new double[normalizedData.length][normalizedData[0].length];
				
		for (int j = 0; j < normalizedData[0].length; j++) {
			for (int i = 0; i < normalizedData.length; i++) {
								
				originalData[i][j] = 
						((normalizedData[i][j]-lowerLimit[j])
						* (upperValue[j] - lowerValue[j]))
						/ (upperLimit[j] - lowerLimit[j])
						+ lowerValue[j];
												
			}

		}
		return originalData;
		
	} //unnormalize()

	/* 
	 * @see neuralnetworktoolkit.normalization.INormalization#setupParameters(double[][])
	 */
	public void setupParameters(double[][] originalData) {
		upperLowerLimitsSetup(originalData);
		upperLowerValuesSetup(originalData);

	} //train()

	/* 
	 * @see neuralnetworktoolkit.normalization.Normalization#transform(double[][])
	 */
	protected double[][] transform(double[][] originalData) {
		double[][] normalizedData;
		
		normalizedData = new double[originalData.length][originalData[0].length];
		
		for (int j = 0; j < originalData[0].length; j++) {
			for (int i = 0; i < originalData.length; i++) {
				normalizedData[i][j] =
					(originalData[i][j] - lowerValue[j])
						* (upperLimit[j] - lowerLimit[j])
						/ (upperValue[j] - lowerValue[j])
						+ lowerLimit[j];

			}

		}
		return normalizedData;
		
	} //transform()
	
	/**
	 * 
	 * @param originalData
	 */
	private void upperLowerLimitsSetup(double[][] originalData) {
		lowerLimit = new double[originalData[0].length];
		upperLimit = new double[originalData[0].length];
		
		for(int i = 0; i < originalData[0].length; i++) {
			lowerLimit[i] = -1;
			upperLimit[i] = 1;
			
		}
		
	} //upperLowerLimitsSetup()
	
	/**
	 * 
	 * @param originalData
	 */
	private void upperLowerValuesSetup(double[][] originalData) {
		lowerValue = new double[originalData[0].length];
		upperValue = new double[originalData[0].length];
		
		for(int i = 0; i < lowerValue.length; i++) {
			lowerValue[i] = Double.MAX_VALUE;
			upperValue[i] = Double.MIN_VALUE;
			
		}
		
		for (int j = 0; j < originalData[0].length; j++) {
			for (int i = 0; i < originalData.length; i++) {
				if (originalData[i][j] > upperValue[j]) {
					upperValue[j] = originalData[i][j];
				}
				if (originalData[i][j] < lowerValue[j]) {
					lowerValue[j] = originalData[i][j];
				}

			}

		}

	} //upperLowerValuesSetup()

	/**
	 * @return Returns the lowerLimit.
	 */
	public double[] getLowerLimit() {
		return lowerLimit;
		
	} //getLowerLimit()
	
	/**
	 * @return Returns the lowerValue.
	 */
	public double[] getLowerValue() {
		return lowerValue;
		
	} //getLowerValue()
	
	/**
	 * @return Returns the upperLimit.
	 */
	public double[] getUpperLimit() {
		return upperLimit;
		
	} //getUpperLimit()
	
	/**
	 * @return Returns the upperValue.
	 */
	public double[] getUpperValue() {
		return upperValue;
		
	} //getUpperValue()
	
} //Linear