-
Notifications
You must be signed in to change notification settings - Fork 0
/
data.js
185 lines (139 loc) · 4.65 KB
/
data.js
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
/*
*
* DATA related functions
*/
// Samples 20 means, 10 for each class for later data generation
var sample_means = function() {
var identidy_mat = [
[ 2.0, 0.0,],
[ 0.0, 2.0]
];
var mean_sample_size = 10;
var mean_distribution_class1 = MultivariateNormal([0.5,-0.5], identidy_mat)
var mean_distribution_class2 = MultivariateNormal([-0.5,0.5], identidy_mat)
var means_class_1 = [];
var means_class_2 = [];
for(var i = 0; i < mean_sample_size; i++)
{
var mean_1 = mean_distribution_class1.sample()
means_class_1.push({
'x': mean_1[0],
'y': mean_1[1],
'label': 1
});
var mean_2 = mean_distribution_class2.sample();
means_class_2.push({
'x': mean_2[0],
'y': mean_2[1],
'label': -1
});
}
return [d3.shuffle(means_class_1), d3.shuffle(means_class_2)]
}
//Samples sampleSize datapoints from the means vector
var generate_mixture_from_means = function(sampleSize, noise, means)
{
var means_class_1 = means[0]
var means_class_2 = means[1]
var data = []
var identidy_mat_small = [
[noise, 0.0,],
[ 0.0, noise]
];
for(var i = 0; i < sampleSize/2; i++)
{
var idx = Math.floor(Math.random()*means_class_1.length)
var mean_class_1 = means_class_1[idx];
var distribution_class_1 = MultivariateNormal([mean_class_1.x, mean_class_1.y], identidy_mat_small)
var sample_point = distribution_class_1.sample()
data.push({
'x': sample_point[0],
'y': sample_point[1],
'label': 1,
'mean': idx
})
var idx2 = Math.floor(Math.random()*means_class_2.length);
var mean_class_2 = means_class_2[idx2];
var distribution_class_2 = MultivariateNormal([mean_class_2.x, mean_class_2.y], identidy_mat_small)
sample_point = distribution_class_2.sample()
data.push({
'x': sample_point[0],
'y': sample_point[1],
'label': -1,
'mean': idx2
})
}
return data
}
//Construct KD-tree for data
var constructKDTree = function(points) {
//Euclidean distance betweeen two points without the square root for optimization
//since we only need relative distances
var distance = function(a, b){
return Math.pow(a.x - b.x, 2) + Math.pow(a.y - b.y, 2);
}
var tree = new kdTree(points, distance, ["x", "y"]);
return tree;
}
var getDartBoardPoints = function(bias, variance) {
var theta = Math.floor(Math.random() * (360 - 0 + 1)) + 0;
var xCoord = bias*math.cos(math.unit(theta, 'deg'));
var yCoord = bias*math.sin(math.unit(theta, 'deg'));
var points = []
var n = 20;
var identidy_mat = [
[ variance, 0.0,],
[ 0.0, variance]
];
var distribution = MultivariateNormal([xCoord, yCoord], identidy_mat)
for(var j = 0; j < 10; j++) {
var point = distribution.sample();
points.push({
'x': point[0],
'y': point[1]
});
}
return points
}
var getVarianceOfModel = function(noise, k)
{
return noise/k;
}
var getBiasOfModel = function(means, k)
{
var num_training_sets = 100;
var errors = []
var meanvector = means[0].concat(means[1])
for(var i = 0; i < num_training_sets; i++)
{
var data = generate_mixture_from_means(400,noiseMap(noise), means);
// filter out points that fall out of our domain
data = data.filter(p => {
return p.x >= domain[0] && p.x <= domain[1]
&& p.y >= domain[0] && p.y <= domain[1];
});
var tree = constructKDTree(data);
var error = calculate_error(tree, meanvector, kval);
errors.push(error)
}
return errors
}
var calculate_error = function(tree,dat, kval) {
var error = 0;
for(var i = 0; i<dat.length; i++)
{
var nearest = tree.nearest({ x: dat[i].x, y: dat[i].y }, kval);
var sum = 0;
var votes = []
for (var h = 0; h < nearest.length; h++)
{
votes.push(nearest[h][0].label)
sum = sum + (nearest[h][0].label == -1 ? -1 : 1);
}
var predicted_label = sum > 0 ? 1 : -1;
//console.log("Predicted/True label: {0}, {1} ".format(predicted_label, dat[i].label))
var error = error + (predicted_label == dat[i].label ? 0 : 1);
// console.log(error)
}
return error/dat.length;
}