Package types

Examples of types.SparseVector

30
31
32
33
34
35
36
37
38
39
40
        SparseVector[] x = new SparseVector[words.size()];
        Object[] y = new Object[tags.size()];
        for (int t = 0; t < x.length; t++) {
          y[t] = tags.get(t);
          String word = "^" + words.get(t) + "$";
          x[t] = new SparseVector();
          x[t].add(xAlphabet.lookupObject(word), 1);
          for (int i = 0; i < word.length() - 3; i++) {
            x[t].add(xAlphabet.lookupObject(word
                .substring(i, i + 3)), 1);
          }
View Full Code Here
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
    ArrayList<ClassificationInstance> result = new ArrayList<ClassificationInstance>();
    BufferedReader reader = new BufferedReader(new FileReader(dataLoc));
    String ln = reader.readLine();
    while (ln != null) {
      String[] feats = ln.split(",");
      SparseVector x = new SparseVector();
      Object y = feats[feats.length - 1];
      for (int i = 3; i < feats.length - 1; i++) {
        if (feats[i].equals("1"))
          x.add(xAlphabet.lookupObject(i), 1.0);
      }
      result.add(new ClassificationInstance(xAlphabet, yAlphabet, x, y));
      ln = reader.readLine();
    }
    return result;
View Full Code Here
38
39
40
41
42
43
44
45
46
47
48
49
50
          for (String w : ln.split("\\s\\s*")) {
            words.add(w);
          }
        }
        reader.close();
        SparseVector x = new SparseVector();
        for (String w : words)
          x.add(xAlphabet.lookupObject(w), 1);
        result.add(new ClassificationInstance(xAlphabet, yAlphabet, x,
            label));
      }
    }
    return result;
View Full Code Here
20
21
22
23
24
25
26
27
28
29
30
31
32
    numY = yAlphabet.size();
    defalutFeatureIndex = xAlphabet.size();
  }

  public SparseVector apply(SparseVector x, int y) {
    SparseVector res = new SparseVector();
    for (int i = 0; i < x.numEntries(); i++) {
      res.add(y * (defalutFeatureIndex + 1) + x.getIndexAt(i), x
          .getValueAt(i));
    }
    res.add(y * (defalutFeatureIndex + 1) + defalutFeatureIndex, 1);
    return res;
  }
View Full Code Here
29
30
31
32
33
34
35
36
37
38
39
40
    this.numInputs = xAlphabet.size() + 1;
  }

  // see SequenceFeatureFunction for java doc
  public SparseVector apply(SparseVector[] x, int[] y) {
    SparseVector result = this.apply(x, 0, y[0], 0);
    for (int t = 1; t < x.length; t++) {
      SparseVector fsubt = this.apply(x, y[t - 1], y[t], t);
      StaticUtils.plusEquals(result, fsubt);
    }
    return result;
  }
View Full Code Here
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
        this.yAsize * this.numInputs;
  }

  // see SequenceFeatureFunction for java doc
  public SparseVector apply(SparseVector[] xseq, int ytm1, int yt, int t) {
    SparseVector result = new SparseVector();
    for (int i = 0; i < xseq[t].numEntries(); i++) {
      int x = xseq[t].getIndexAt(i);
      // initial yt
      if (t == 0) {
        result.add(this.initialIndex(yt, x), 1);
        // ymt1 times yt
      } else {
        result.add(this.labelXlabel(ytm1, yt, x), 1);
      }
      if (t == xseq.length - 1) {
        result.add(this.finalIndex(yt, x), 1);
      }
      // yt times x
      result.add(this.labelXinput(yt, x), 1);
    }
    int x = this.xAsize;
    // initial yt
    if (t == 0) {
      result.add(this.initialIndex(yt, x), 1);
      // ymt1 times yt
    } else {
      result.add(this.labelXlabel(ytm1, yt, x), 1);
    }
    if (t == xseq.length - 1) {
      result.add(this.finalIndex(yt, x), 1);
    }
    // yt times x
    result.add(this.labelXinput(yt, x), 1);

    return result;
  }
View Full Code Here
18
19
20
21
22
23
24
25
26
27
28
29
    this.yAsize = yAlphabet.size();
  }

  // see SequenceFeatureFunction for java doc
  public SparseVector apply(SparseVector[] x, int[] y) {
    SparseVector result = apply(x, 0, y[0], 0);
    for (int t = 1; t < x.length; t++) {
      SparseVector fsubt = apply(x, y[t - 1], y[t], t);
      StaticUtils.plusEquals(result, fsubt);
    }
    return result;
  }
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
        yAsize * xAsize;
  }

  // see SequenceFeatureFunction for java doc
  public SparseVector apply(SparseVector[] x, int ytm1, int yt, int t) {
    SparseVector result = new SparseVector();
    // initial yt
    if (t == 0)
      result.add(initialIndex(yt), 1);
    // ymt1 times yt
    else
      result.add(labelXlabel(ytm1, yt), 1);
    if (t == x.length - 1)
      result.add(finalIndex(yt), 1);
    // yt times x
    for (int i = 0; i < x[t].numEntries(); i++) {
      result.add(labelXinput(yt, x[t].getIndexAt(i)), 1);
    }
    return result;
  }
View Full Code Here
82
83
84
85
86
87
88
89
90
91
92
93
94
    double wrongUpdate = Math.exp(alpha);
    double correctUpdate = Math.exp(-alpha);
    for (int instInd = 0; instInd < trainingData.size(); instInd++) {
      ClassificationInstance inst = trainingData.get(instInd);
      for (int y = 0; y < yAlphabet.size(); y++) {
        SparseVector fv = fxy.apply(inst.x, y);
        for (int i = 0; i < fv.numEntries(); i++) {
          if (fv.getIndexAt(i) == bestFeature) {
            if (y == inst.y)
              w[instInd] *= correctUpdate;
            else
              w[instInd] *= wrongUpdate;
          }
View Full Code Here
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
    }
    for (int instInd = 0; instInd < trainingData.size(); instInd++) {
      ClassificationInstance inst = trainingData.get(instInd);
      total += w[instInd];
      for (int y = 0; y < yAlphabet.size(); y++) {
        SparseVector fv = fxy.apply(inst.x, y);
        if (y == inst.y) {
          for (int i = 0; i < fv.numEntries(); i++) {
            correct[fv.getIndexAt(i)] += w[instInd];
          }
        } else {
          for (int i = 0; i < fv.numEntries(); i++) {
            wrongs[fv.getIndexAt(i)] += w[instInd];
          }
        }
      }
    }
    for (int i = 0; i < correct.length; i++) {
View Full Code Here
TOP

Related Classes of types.SparseVector

  • classification.AdaBoost
  • classification.CompleteFeatureFunction
  • gnu.trove.TDoubleArrayList
  • gnu.trove.TIntArrayList
  • io.InternetAdReader
  • io.NewsgroupsReader
  • io.PartOfSpeechReader
  • sequence.OneYwithXFeatureFunction
  • sequence.TwoYwithXFeatureFunction
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.