package algorithm;
import org.encog.Encog;
import org.encog.ml.data.MLData;
import org.encog.ml.data.MLDataPair;
import org.encog.ml.data.MLDataSet;
import org.encog.ml.data.basic.BasicMLDataSet;
import org.encog.ml.svm.KernelType;
import org.encog.ml.svm.SVMType;
import org.encog.ml.svm.training.SVMTrain;
import org.encog.neural.NeuralNetworkError;
import java.util.ArrayList;
/**
* Created by fish123 on 2016/12/14.
*/
public class SVM {
org.encog.ml.svm.SVM svm;
public void trainSVM(double[][] trainData,double[][] trainlabel,int SVMtype,int kenelType){
org.encog.ml.svm.SVMType svm_type;
switch(SVMtype) {
case 1:
svm_type = SVMType.NewSupportVectorRegression;
break;
case 2:
svm_type = SVMType.NewSupportVectorClassification;
break;
case 3:
svm_type = SVMType.SupportVectorOneClass;
break;
case 4:
svm_type = SVMType.EpsilonSupportVectorRegression;
break;
case 5:
svm_type = SVMType.NewSupportVectorRegression;
break;
default:
throw new NeuralNetworkError("Invalid svm type");
}
org.encog.ml.svm.KernelType kernel_type;
switch(kenelType) {
case 1:
kernel_type = KernelType.Linear;
break;
case 2:
kernel_type = KernelType.Poly;
break;
case 3:
kernel_type = KernelType.RadialBasisFunction;
break;
case 4:
kernel_type = KernelType.Sigmoid;
break;
case 5:
kernel_type = KernelType.Precomputed;
break;
default:
throw new NeuralNetworkError("Invalid svm type");
}
svm = new org.encog.ml.svm.SVM(trainData[0].length,svm_type, kernel_type);
// create training data
MLDataSet trainingSet = new BasicMLDataSet(trainData,trainlabel);
// train the SVM
final SVMTrain train = new SVMTrain(svm, trainingSet);
train.iteration();
train.finishTraining();
}
public Double[][] computeSVM(double[][] testData)
{
double[][] dataTotest = new double[testData.length][testData[0].length];
double[][] TLabel = new double[testData.length][1];
for(int i=0;i<dataTotest.length;i++){
TLabel[i][0] = Math.random();
}
MLDataSet testSet = new BasicMLDataSet(dataTotest,TLabel);
// test the SVM
System.out.println("SVM Results:");
int index = 0;
ArrayList<Double> result = new ArrayList<>();
String content = "";
for(MLDataPair pair: testSet ) {
final MLData output = svm.compute(pair.getInput());
content+= "predist=" + output.getData(0)+"\r\n";
result.add(output.getData(0));
index++;
}
Double[][] returnResult = new Double[result.size()][1];
for(int i = 0;i<result.size();i++)
{
returnResult[i][0] = result.get(i);
}
return returnResult;
}
}
