-
Notifications
You must be signed in to change notification settings - Fork 3
/
experiment_rbm_sumstats.m
52 lines (39 loc) · 1.75 KB
/
experiment_rbm_sumstats.m
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
% Experiment with rbm features and voting classification
addpath('Voicebox');
addpath('rbm');
num_features = 300;
fprintf('Building trainingset...\n');
[trainingSet, trainingLabels, w_training] = getDatasetFromDir('data/training',0);
fprintf('Building and training Restricted Boltzmann Machine...\n');
rbm = build_rbm(trainingSet, num_features);
fprintf('Building testset...\n');
[testSet, testLabels, w_test] = getDatasetFromDir('data/test',0);
fprintf('Extracting features...\n');
training = zeros(size(trainingSet,1), num_features);
for i = 1:size(trainingSet,1)
training(i,:) = rbm_get_hidden(trainingSet(i,:), rbm);
end
testing = zeros(size(testSet,1), num_features);
for i = 1:size(testSet,1)
testing(i,:) = rbm_get_hidden(testSet(i,:), rbm);
end
training_sumstats = zeros(length(w_training), num_features * 2);
cum = cumsum([0; w_training]);
training_labels = zeros(length(w_training), 1);
for i = 1:length(w_training)
training_sumstats(i,:) = [mean(training(cum(i) + 1: cum(i + 1),:)) std(training(cum(i) + 1: cum(i + 1),:))];
training_labels(i) = trainingLabels(cum(i + 1));
end
test_sumstats = zeros(length(w_test), num_features * 2);
cum = cumsum([0; w_test]);
test_labels = zeros(length(w_training), 1);
for i = 1:length(w_test)
test_sumstats(i,:) = [mean(testing(cum(i) + 1: cum(i + 1),:)) std(testing(cum(i) + 1: cum(i + 1),:))];
test_labels(i) = testLabels(cum(i + 1));
end
fprintf('Training...\n');
model = svmtrain(training_labels, training_sumstats);
fprintf('Testing...\n');
[predictedLabels, accuracy] = svmpredict(test_labels, test_sumstats, model);
performance = sum(predictedLabels==test_labels)/length(test_labels);
fprintf('Performance: %2.4f (%i/%i)\n', performance, sum(predictedLabels==test_labels), length(test_labels));