MachineLearning/Whitening.java at master · CodeMyWorld/MachineLearning · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
import org.apache.commons.math3.linear.Array2DRowRealMatrix;
import org.apache.commons.math3.linear.EigenDecomposition;
import org.apache.commons.math3.linear.RealMatrix;
import org.apache.commons.math3.stat.correlation.Covariance;
import java.util.ArrayList;

/**
 * Created by alex on 15-5-30.
 */
public class Whitening {
    private DataBuilder dataBuilder;

    public Whitening(DataBuilder dataBuilder){
        this.dataBuilder = dataBuilder;
    }

    public void PCAWhitening(ArrayList<ArrayList<Double>> data){
        PreTreatment preTreatment = new PreTreatment(dataBuilder);
        preTreatment.SubstracMean(dataBuilder.getData());
        preTreatment.NormalizeVariance(dataBuilder.getData());

        RealMatrix dataMatrix = new Array2DRowRealMatrix(dataBuilder.getRow(),dataBuilder.getColumn());

        for(int i = 0; i < data.size(); i++){
            for(int j = 0; j < data.get(i).size(); j++){
                dataMatrix.addToEntry(i,j,data.get(i).get(j));
            }
        }

        Covariance covariance = new Covariance(dataMatrix);
        RealMatrix covarianceMatrix = covariance.getCovarianceMatrix();
        EigenDecomposition eigenDecomposition = new EigenDecomposition(covarianceMatrix);
        RealMatrix U = eigenDecomposition.getV();
        double eigenvalues[] = eigenDecomposition.getRealEigenvalues();
        U = U.transpose();
        for(int i = 0; i < dataBuilder.getRow(); i++){
            dataMatrix.setRowVector(i,U.operate(dataMatrix.getRowVector(i)));
        }

        Covariance covariance2 = new Covariance(dataMatrix);
        RealMatrix result = covariance2.getCovarianceMatrix();


        for(int i = 0; i < dataMatrix.getColumnDimension(); i++){

            dataMatrix.setColumnVector(i, dataMatrix.getColumnVector(i).mapDivideToSelf(Math.sqrt(result.getEntry(i, i) + 0.00001)));
        }

        Covariance covariance3 = new Covariance(dataMatrix);
        RealMatrix finalResult = covariance3.getCovarianceMatrix();

        for(int i = 0; i < 100; i++){
            for(int j = 0; j < 100; j++){
                System.out.print(finalResult.getEntry(i, j) + " ");
            }
            System.out.println();
        }
    }

    private RealMatrix[] PCAWhiteningForZCAWhitening(ArrayList<ArrayList<Double>> data){
        PreTreatment preTreatment = new PreTreatment(dataBuilder);
        preTreatment.SubstracMean(dataBuilder.getData());
        preTreatment.NormalizeVariance(dataBuilder.getData());

        RealMatrix dataMatrix = new Array2DRowRealMatrix(dataBuilder.getRow(),dataBuilder.getColumn());


        for(int i = 0; i < data.size(); i++){
            for(int j = 0; j < data.get(i).size(); j++){
                dataMatrix.addToEntry(i,j,data.get(i).get(j));
            }
        }

        Covariance covariance = new Covariance(dataMatrix);
        RealMatrix covarianceMatrix = covariance.getCovarianceMatrix();

        System.out.println("covariance matrix of original data after zero-mean process:");
        for(int i = 0; i < 2; i++){
            for(int j = 0; j < 2; j++){
                System.out.print(covarianceMatrix.getEntry(i,j) + " ");
            }
            System.out.println();
        }

        EigenDecomposition eigenDecomposition = new EigenDecomposition(covarianceMatrix);
        RealMatrix U = eigenDecomposition.getV();
        U = U.transpose();
        for(int i = 0; i < dataBuilder.getRow(); i++){
            dataMatrix.setRowVector(i,U.operate(dataMatrix.getRowVector(i)));
        }

        Covariance covariance2 = new Covariance(dataMatrix);
        RealMatrix result = covariance2.getCovarianceMatrix();


        for(int i = 0; i < dataMatrix.getColumnDimension(); i++){

            dataMatrix.setColumnVector(i, dataMatrix.getColumnVector(i).mapDivideToSelf(Math.sqrt(result.getEntry(i, i) + 0.00001)));
        }

        Covariance covariance3 = new Covariance(dataMatrix);
        RealMatrix finalResult = covariance3.getCovarianceMatrix();

        System.out.println("\nthe covariance matrix of data after PCA whitening:");
        for(int i = 0; i < 2; i++){
            for(int j = 0; j < 2; j++){
                System.out.print(finalResult.getEntry(i, j) + " ");
            }
            System.out.println();
        }
        RealMatrix returnRealMatrix[] = new RealMatrix[2];
        returnRealMatrix[0] = eigenDecomposition.getV();
        returnRealMatrix[1] = dataMatrix;
        return returnRealMatrix;
    }

    public void ZCAWhitening(ArrayList<ArrayList<Double>> data){
        RealMatrix[] uAndDataMatrix = PCAWhiteningForZCAWhitening(data);
        RealMatrix U = uAndDataMatrix[0];
        RealMatrix dataMatrix = uAndDataMatrix[1];
        for(int i = 0; i < dataMatrix.getRowDimension(); i++){
            dataMatrix.setRowVector(i,U.operate(dataMatrix.getRowVector(i)));
        }

        Covariance covariance3 = new Covariance(dataMatrix);
        RealMatrix finalResult = covariance3.getCovarianceMatrix();

        System.out.println("\nthe covariance matrix of data after ZCA whitening:");
        for(int i = 0; i < 2; i++){
            for(int j = 0; j < 2; j++){
                System.out.print(finalResult.getEntry(i,j) + " ");
            }
            System.out.println();
        }
    }
}