forked from DraXus/online-shutdown-identification
-
Notifications
You must be signed in to change notification settings - Fork 0
/
extra_functions.m
54 lines (49 loc) · 1.07 KB
/
extra_functions.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
53
54
%data = normalizedData(1:200000,5);
data = data_generator(1);
window = 50;
rows = size(data,1);
newdata = zeros(1,rows);
% Sliding window mean
for i=1:rows
if(i<=window)
newdata(i) = data(i);
else
newdata(i) = mean(data(i-window:i));
end
end
% Cumulative mean
for i=1:rows
newdata(i) = mean(data(1:i));
end
% Log-likelihood Gaussian Mixture Model
obj = gmdistribution.fit(data,2);
m0 = obj.mu(1);
m1 = obj.mu(2);
s0 = obj.Sigma(1);
s1 = obj.Sigma(2);
S = zeros(1,rows);
for i=1:rows
p0 = pdf('Normal', data(i), m0, s0);
p1 = pdf('Normal', data(i), m1, s1);
S(i) = log(p0/p1);
if(i<=window)
newdata(i) = S(i);
else
newdata(i) = sum(S(i-window:i));
end
end
% Log-likelihood (windowed)
S = zeros(1,rows);
m0 = median(data);
s0 = mad(data);
for i=1:rows
if(i<=window)
newdata(i) = 0; % equiprobable
else
%m0 = mean(data(i-window:i));
%s0 = std(data(i-window:i));
p0 = pdf('Normal', data(i), m0, s0);
S(i) = log(p0/(1-p0));
newdata(i) = sum(S(i-window:i));
end
end