dbtest.java

利用Java实现的神经网络工具箱

/*
 * NetworkController.java	1.0  30/10/2004
 *
 * 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.tests;

import neuralnetworktoolkit.NetworkController;
import neuralnetworktoolkit.NetworkControllerException;
import neuralnetworktoolkit.NeuralModel;
import neuralnetworktoolkit.StatisticalResults;
import neuralnetworktoolkit.datamanager.databasemanager.DbaseManager;
import neuralnetworktoolkit.datamanager.textfilemanager.LoadFromFileManager;
import neuralnetworktoolkit.math.NeuralMath;
import neuralnetworktoolkit.methods.gradientbased.quasinewton.lm.*;
import neuralnetworktoolkit.validation.ValidationStatistics;

/**
 * @author iver
 *
 * TODO To change the template for this generated type comment go to
 * Window - Preferences - Java - Code Generation - Code and Comments
 */
public class DBTest {
	
	public static void main(String[] args) {
		
		LoadFromFileManager lf = new LoadFromFileManager();
		DbaseManager db = new DbaseManager("postgresql");
		//db.setDriverName("org.postgresql.Driver");
		db.connectDatabase("jdbc:postgresql://localhost/Iver","iver","x");
		//db.setDriverName("com.mysql.jdbc.Driver");
		//db.connectDatabase("jdbc:mysql://localhost/teste");
		
		String[] tables = db.getTableNames();
		String[] columns = null;
		
		System.out.println("Tabelas Existentes:");
		for (int i = 0; i < tables.length; i++) {
			System.out.println("Nome da Tabela: " + tables[i]);
			columns = db.getTableColumnNames(tables[i]);
			for (int j = 0; j < columns.length; j++) {
				System.out.println("Nome da Coluna: " + columns[j]);
			}
			System.out.println();
		}
		String[] trC = {"a","b"}; 
		double[][] data = db.loadData("teste", trC);
		NeuralMath.printMatrix(data);
						
		StatisticalResults statistics;
		ValidationStatistics validationStatistics;
		
		System.out.println("Teste iniciado!");
		
		NetworkController nc = new NetworkController();
		
		try {
						
			int[] parameters = {1, 10, 1};
			
			String[] functions = {"Tanh", "Tanh"};
			
			String initialization = "NguyenWidrow";
			nc.createNeuralNetwork(parameters, initialization, functions);
			
			nc.setNormalization("Linear", NetworkController.FULLNORMALIZER);
						
			int[] index = {1};
			nc.setOutputIndex(index);
						
			nc.setTrainingData(data);
			nc.splitDataValues();
			
			nc.setTrainingMethod("gradientbased.quasinewton.lm.OLMAM");
						
			LevenbergMarquardtParameters trainParam = new LevenbergMarquardtParameters();
			
			trainParam.setError(.000000000000000000000000005);
			trainParam.setLmKind(LevenbergMarquardtModel.INCOMPLETE_LM);
			trainParam.setMaxIterations(10);
						
			statistics = nc.trainNeuralNetwork(trainParam);
			
			System.out.println("Número de épocas: " + statistics.getNumberOfEpochs());
			System.out.println("Número de iterações: " + statistics.getNumberIterations());
			System.out.println("Erro do treinamento: " + statistics.getError());
			System.out.println("Tempo de treinamento: " + statistics.getTrainingTime());
			
			String[] inC = {"c"};
			double[][] tdata = db.loadData("teste", inC);
			System.out.println("-----> Inferencia <-----");
			nc.setInferenceData(tdata);
			nc.infereWithNetwork();
			System.out.println("-----> Inferencia <-----");
			
			db.saveNeuralModel("blo","hugo", nc.getNeuralModel());
			
		} catch (NetworkControllerException e) {
			e.printStackTrace();
		}
		
		System.out.println("Teste finalizado!");		
				
		String[][] existentModels;
		existentModels = db.getExistitentModels("blo", "hugo",.0005,null);
		
		if (existentModels != null) {
			System.out.println("Modelos existentes:");
			for (int i = 0; i < existentModels.length; i++) {
				for (int j = 0; j < existentModels[0].length; j++) {
					System.out.print(existentModels[i][j] + " ");
				}
				System.out.println();
			}
		}
		
		String[] chave = {"hugo","2004-11-14", "02:04:45"};
		
		NetworkController retrievedNN = new NetworkController();
		try {
			
			NeuralModel nm = db.retrieveModel("blo", chave);
			if (nm != null ) {
				System.out.println("Recuperei nao é nulo");
				retrievedNN.createNeuralNetwork(nm);
				String[] inC = {"c"};
				double[][] tdata = db.loadData("teste", inC);
				retrievedNN.setInferenceData(tdata);
				System.out.println("-----> Inferencia <-----");
				retrievedNN.infereWithNetwork();
				System.out.println("-----> Inferencia <-----");
			}
			
		} catch (NetworkControllerException e1) {
			e1.printStackTrace();
		}
		
		db.disconnectDatabase();
		
	} // main()
	
}