Machine-Learning/visualize_Logistic_Regression.html at master · Xelio/Machine-Learning · 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
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
<html>
<head>
</head>
<body>
<script src="http://code.jquery.com/jquery-1.10.1.min.js"></script>
<script type="text/JavaScript" src="./js/lib/jsDraw2D.js"></script>
<script type="text/JavaScript" src="./js/drawing.js"></script>
<script type="text/JavaScript" src="./js/lib/math.min.js"></script>
<script type="text/JavaScript" src="./js/TestPlane.js"></script>
<script type="text/JavaScript" src="./js/Logistic_Regression.js"></script>
<div id="control">
  <div>
    <label for="N">N = </label>
    <input type="text" name="N" id="N" value="100" size="10" />
    <label for="testTimes">Test times = </label>
    <input type="text" name="testTimes" id="testTimes" value="100" size="10" />
    <label for="eta">eta = </label>
    <input type="text" name="eta" id="eta" value="0.01" size="10" />
  </div>
  <br />
  <div>
    <button id="lr_anim">Animate Logistic Regression(by point)</button>
  </div>
  <div>
    <button id="lr_anim_epoch">Animate Logistic Regression(by epoch)</button>
  </div>
  <div>
    <button id="lr_test_times">Test Logistic Regression Multiple Times</button>
  </div>
  <div>
    <button id="stop">Stop</button>
  </div>
</div>
<div style="position:relative;">
  <div style="position:absolute; width: 400px;">
    <textarea id="status" readonly rows="30" cols="40"></textarea>
  </div>
  <div id="canvas" style="position:absolute; width:500px;height:500px;left:400px" />
</div>
<script type="text/JavaScript">
  var numberOfErrorTestPoint = 1000;

  var add_w0 = function(point) {
    return [1,
            point[0],
            point[1]];
  };

  window.generateDataSet = function(NumberOfN, plane) {
    var result = [];
    for(var i = 0; i < NumberOfN; i++) {
      var point = plane.randomPoint();
      var output = plane.getOutput(point);
      result.push({point: point, output: output});
    }
    return result;
  }

  window.transformDataSet = function(dataSet, transformFunction) {
    var result = [];
    for(var i in dataSet) {
      result.push({point: transformFunction(dataSet[i].point), output: dataSet[i].output});
    }
    return result;
  }


  window.animateLogisticRegression = function(NumberOfN, eta) {
    var plane = new window.TestPlane(-1,1,-1,1);
    var dataSet = window.generateDataSet(NumberOfN, plane);
    var dataSetWithw0 = window.transformDataSet(dataSet, add_w0);

    var logisticRegression = new window.Logistic_Regression();
    logisticRegression.initializeLearning(dataSetWithw0, [0,0,0], eta, 0.01);

    var completeCallback = function(norm_history, weights_history, epoch) {
      window.initCanvas();
      window.drawTargetLine(plane)
      window.drawDataSet(dataSet);
      window.drawRegressionResult(weights_history[weights_history.length -1], plane);

      var EoutDataSet = window.transformDataSet(window.generateDataSet(numberOfErrorTestPoint, plane), add_w0);
      var Eout = logisticRegression.Eout(EoutDataSet);
      window.displayResult(norm_history, weights_history, epoch, Eout);
    };

    var stepCallback = function(data, norm, weights, norm_history, weights_history, epoch, remaining_points) {
      window.initCanvas();
      window.drawTargetLine(plane)
      window.drawDataSet(dataSet);
      window.drawRegressionStatus(weights, weights_history, data.point, plane);

      window.displayStatus(norm, weights, norm_history[norm_history.length - 1], weights_history[weights_history.length - 1], epoch, remaining_points);
      window.timer = setTimeout(function() {logisticRegression.stochasticGradientDescentStep(stepCallback, null, completeCallback);}, 50);
    };

    window.initCanvas();
    logisticRegression.stochasticGradientDescentStep(stepCallback, null, completeCallback);
  }

  window.animateLogisticRegressionEpoch = function(NumberOfN, eta) {
    var plane = new window.TestPlane(-1,1,-1,1);
    var dataSet = window.generateDataSet(NumberOfN, plane);
    var dataSetWithw0 = window.transformDataSet(dataSet, add_w0);

    var logisticRegression = new window.Logistic_Regression();
    logisticRegression.initializeLearning(dataSetWithw0, [0,0,0], eta, 0.01);

    var completeCallback = function(norm_history, weights_history, epoch) {
      window.initCanvas();
      window.drawTargetLine(plane)
      window.drawDataSet(dataSet);
      window.drawRegressionResult(weights_history[weights_history.length -1], plane);

      var EoutDataSet = window.transformDataSet(window.generateDataSet(numberOfErrorTestPoint, plane), add_w0);
      var Eout = logisticRegression.Eout(EoutDataSet);
      window.displayResult(norm_history, weights_history, epoch, Eout);
    };

    var epochCallback = function(norm_history, weights_history, epoch) {
      window.initCanvas();
      window.drawTargetLine(plane)
      window.drawDataSet(dataSet);
      window.drawRegressionEpoch(weights_history, plane);

      window.displayEpoch(norm_history, weights_history, epoch);
      window.timer = setTimeout(function() {logisticRegression.stochasticGradientDescentEpochStep(epochCallback, completeCallback);}, 50);
    };

    window.initCanvas();
    logisticRegression.stochasticGradientDescentEpochStep(epochCallback, completeCallback);
  }

  window.testLogisticRegressionMultipleTimes = function(NumberOfN, eta, times) {
    var logisticRegression = new window.Logistic_Regression();

    var totalEout = 0;
    var totalEpoch = 0;
    var currentTimes = 0;
    var plane;
    var newTest;
    var completeCallback;
    var dataSet;
    var dataSetWithw0;

    newTest = function() {
      plane = new window.TestPlane(-1,1,-1,1);
      dataSet = window.generateDataSet(NumberOfN, plane);
      dataSetWithw0 = window.transformDataSet(dataSet, add_w0);
      logisticRegression.initializeLearning(dataSetWithw0, [0,0,0], eta, 0.01);
      logisticRegression.stochasticGradientDescent(completeCallback);
    }

    completeCallback = function(norm_history, weights_history, epoch) {
      window.initCanvas();
      window.drawTargetLine(plane)
      window.drawDataSet(dataSet);
      window.drawRegressionResult(weights_history[weights_history.length -1], plane);

      var EoutDataSet = window.transformDataSet(window.generateDataSet(numberOfErrorTestPoint, plane), add_w0);
      var Eout = logisticRegression.Eout(EoutDataSet);

      totalEout += Eout;
      totalEpoch += epoch;
      currentTimes += 1;

      if(currentTimes < times) {
        var status = 'Test: ' + currentTimes + '\r\n'
                   + 'epoch: ' + epoch + '\r\n'
                   + 'Eout: ' + Eout;
        $('#status').val(status);

        window.timer = setTimeout(function() {newTest();}, 1);
      } else {
        var avgEout = totalEout / times;
        var avgEpoch = totalEpoch / times;
        var status = 'Average epoch: ' + avgEpoch + '\r\n'
                   + 'Average Eout: ' + avgEout;
        $('#status').val(status);
      }
    };

    window.initCanvas();
    newTest();
  }


  window.displayStatus = function(norm, current_weights, previous_norm, previous_weights, epoch, remaining_points) {
    var status = 'epoch: ' + epoch + '\r\n\r\n'
               + 'weights of previous epoch: ' + '\r\n' + window.joinWithRounding(previous_weights, 6) + '\r\n'
               + 'norm of previous epoch: ' + '\r\n' + previous_norm + '\r\n\r\n'
               + 'weights: ' + '\r\n' + window.joinWithRounding(current_weights, 6) + '\r\n'
               + 'norm: ' + '\r\n' + norm + '\r\n\r\n'
               + 'Remaining points of this epoch: ' + remaining_points;
    $('#status').val(status);
  }

  window.displayEpoch = function(norm_history, weights_history, epoch) {
    var current_weights = weights_history[weights_history.length - 1];
    var previous_weights = weights_history[weights_history.length - 2];

    var status = 'epoch: ' + epoch + '\r\n\r\n'
               + 'weights: ' + '\r\n' + window.joinWithRounding(current_weights, 6) + '\r\n\r\n'
               + 'weights of previous epoch: ' + '\r\n' + window.joinWithRounding(previous_weights, 6) + '\r\n\r\n'
               + 'norm: ' + norm_history[norm_history.length - 1];
    $('#status').val(status);
  }

  window.displayResult = function(norm_history, weights_history, epoch, Eout) {
    var result_weights = weights_history[weights_history.length -1];

    var last_5_weights = '';
    var last_5_norm = '';

    for(var i = 0; i < Math.min(5, weights_history.length - 2); i++) {
      last_5_weights += window.joinWithRounding(weights_history[weights_history.length - 2 - i], 6) + '\r\n';
      last_5_norm += norm_history[norm_history.length - 2 - i] + '\r\n';
    }

    var status = 'total epoch: ' + epoch + '\r\n'
               + 'Eout: ' + Eout + '\r\n'
               + 'weights: ' + '\r\n' + window.joinWithRounding(result_weights, 6) + '\r\n\r\n'
               + 'norm: ' + norm_history[norm_history.length - 1] + '\r\n\r\n'
               + 'last 5 weights: ' + '\r\n' + last_5_weights + '\r\n\r\n'
               + 'last 5 norm: ' + '\r\n' + last_5_norm + '\r\n\r\n';

    $('#status').val(status);
  }

  window.drawRegressionStatus = function(current_weights, weights_history, training_point, plane) {
    if(weights_history.length > 1) {
      var secondLastEpochWeightsLine = window.weightsToLine(weights_history[weights_history.length -2], plane);
      window.plotLine(secondLastEpochWeightsLine, pinkPen);
    }

    var lastEpochWeightsLine = window.weightsToLine(weights_history[weights_history.length - 1], plane);
    window.plotLine(lastEpochWeightsLine, pinkPen2);

    var currentWeightsLine = window.weightsToLine(current_weights, plane);
    window.plotLine(currentWeightsLine, greenPen);

    window.plotCircle([training_point[1], training_point[2]], orangePen);
  }

  window.drawRegressionEpoch = function(weights_history, plane) {
    if(weights_history.length > 2) {
      var secondLastEpochWeightsLine = window.weightsToLine(weights_history[weights_history.length -3], plane);
      window.plotLine(secondLastEpochWeightsLine, pinkPen);
    }

    if(weights_history.length > 1) {
      var lastEpochWeightsLine = window.weightsToLine(weights_history[weights_history.length - 2], plane);
      window.plotLine(lastEpochWeightsLine, pinkPen2);
    }

    var currentWeightsLine = window.weightsToLine(weights_history[weights_history.length - 1], plane);
    window.plotLine(currentWeightsLine, greenPen);
  }

  window.drawRegressionResult = function(result_weights, plane) {
    var resultWeightsLine = window.weightsToLine(result_weights, plane);
    window.plotLine(resultWeightsLine, greenPen);

  }

  window.stop = function() {
    clearTimeout(window.timer);
  }

  window.initCanvas();
  $('#lr_anim').on('click', function() { window.animateLogisticRegression($('#N').val(), parseFloat($('#eta').val())); });
  $('#lr_anim_epoch').on('click', function() { window.animateLogisticRegressionEpoch($('#N').val(), parseFloat($('#eta').val())); });
  $('#lr_test_times').on('click', function() { window.testLogisticRegressionMultipleTimes($('#N').val(), parseFloat($('#eta').val()), $('#testTimes').val()); });

  $('#stop').on('click', function() { window.stop(); });
</script>