/
create_svm.m
168 lines (131 loc) · 4.88 KB
/
create_svm.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
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
function [best_result,flabels] = create_svm()
bins = 0;
removal = 'false';
threshold = 70;
minmax = 1;
reduce = 3;
VOCinit;
[suitability, im_accumulators, rows, ent_results, per_results, classes_cat, c_mu, a_mu, a_cov, c_cov,results, full_results, ful_results,norm_results,norm_results_s] = analyze_data_for_bn(bins, removal, threshold, minmax,reduce);
results = norm_results_s;
[r,c] = size(results);
findex = find(strcmp(results(1,:), 'brightness'));
ind_att = find(strcmp(results(1,:), 'Area'));
%prepare data
f_data = results(2:end, findex:(ind_att-1));
att_data = results(2:end, ind_att:end);
for h = 1:size(att_data,1)
for l = 1: size(att_data, 2)
t{h, l} = mean(cell2mat(f_data(:, l)));
end
end
att_data = t;
att_data(1:2, 1:5)
sel_data = results(2:end,7);
mid_data = cell(r-1,2);
data = [f_data att_data mid_data sel_data];
data_no_att = [f_data mid_data sel_data];
size(data)
size(data_no_att)
[rows coloms] = size(data)
data_ones = [];
data_ones_no_att = [];
for m = 1: rows
if cell2mat(data(m, end))==1
data_ones = [data_ones; data(m, :)];
data_ones_no_att = [data_ones_no_att; data_no_att(m, :)];
end
end
% calculate oroportion of 1 and 2 in sel_data
[temp777 temp222] = hist(cell2mat(sel_data), unique(cell2mat(sel_data)));
temp777
%
difference = temp777(2) - temp777(1)
for l = 1:difference
coef = randi(temp777(1));
data = [data; data_ones(coef, :)];
data_no_att = [data_no_att; data_ones_no_att(coef, :)];
end
sel_data_inc = data(:, end);
sel_data_inc_no_att = data_no_att(:, end);
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
B=cell2mat(data);
B_no_att = cell2mat(data_no_att);
B(:,end)=[];
B_no_att(:, end) = [];
R=cov(B);
[U,D]=eig(R);
B=B*U;
R_na=cov(B_no_att);
[U_na,D_na]=eig(R_na);
B_no_att=B_no_att*U_na;
cvpart = cvpartition(cell2mat(sel_data_inc),'holdout',0.2);
Xtrain = B(training(cvpart),(end-3):(end-1));
Ytrain = sel_data_inc(training(cvpart),:);
Xtest = B(test(cvpart),(end-3):(end-1));
Ytest = sel_data_inc(test(cvpart),:);
model = svmtrain((Xtrain),cell2mat(Ytrain),'kernel_function', 'rbf', 'rbf_sigma', 0.001);
predicted1 = svmclassify(model, (Xtrain));
errRate1 = sum(cell2mat(Ytrain)~= predicted1)/size(cell2mat(Ytrain),1)
predicted = svmclassify(model, (Xtest));
errRate = sum(cell2mat(Ytest)~= predicted)/size(cell2mat(Ytest),1)
cvpart_1 = cvpartition(cell2mat(sel_data_inc_no_att),'holdout',0.2);
Xtrain_1 = B_no_att(training(cvpart_1),(end-3):(end-1));
Ytrain_1 = sel_data_inc_no_att(training(cvpart_1),:);
Xtest = B_no_att(test(cvpart_1),(end-3):(end-1));
Ytest = sel_data_inc_no_att(test(cvpart_1),:);
model = svmtrain((Xtrain_1),cell2mat(Ytrain_1),'kernel_function', 'rbf', 'rbf_sigma', 0.001);
predicted1 = svmclassify(model, (Xtrain_1));
errRate2 = sum(cell2mat(Ytrain_1)~= predicted1)/size(cell2mat(Ytrain_1),1)
predicted2 = svmclassify(model, (Xtest));
errRate3 = sum(cell2mat(Ytest)~= predicted2)/size(cell2mat(Ytest),1)
% % % cvpart = cvpartition(cell2mat(sel_data_inc),'holdout',0.5);
% % % Xtrain = data(training(cvpart),1:end-1);
% % % Ytrain = sel_data_inc(training(cvpart),:);
% % % Xtest = data(test(cvpart),1:end-1);
% % % size(Xtrain)
% % % size(data)
% % % Ytest = sel_data_inc(test(cvpart),:);
% B=cell2mat(data);
% B(:,end)=[];
%
% R=cov(B);
% [U,D]=eig(R);
%
% B=B*U;
%
% t=cell2mat(sel_data_inc);
%
% I=find(t==1);
%
% figure(5)
% plot(B(:,end-1),B(:,end-2),'g.')
% hold
% plot(B(I,end-1),B(I,end-2),'.')
% XX=cell2mat(Xtrain);
% YY=cell2mat(Ytrain);
%learn from all features
%Matlab Svm
% model = svmtrain(cell2mat(Xtrain),cell2mat(Ytrain),'kernel_function', 'rbf', 'rbf_sigma', 1);
% predicted1 = svmclassify(model, cell2mat(Xtrain));
% errRate1 = sum(cell2mat(Ytrain)~= predicted1)/size(cell2mat(Ytrain),1)
% predicted = svmclassify(model, cell2mat(Xtest));
% errRate = sum(cell2mat(Ytest)~= predicted)/size(cell2mat(Ytest),1)
% %libsvm
% model = svmtrain(Ytrain,Xtrain, '-s 1 -c 0.5 -n 0.0001');
% predicted = svmpredict(Ytest, Xtest, model);
%Matlab Boosting
% t = ClassificationTree.template('minleaf',5);
% ens = fitensemble(Xtrain,Ytrain,'AdaBoostM1',100,t,'LearnRate',0.1,'nprint',100);
% predicted = predict(ens,Xtest);
% errRate = sum(Ytest~= predicted)/size(Ytest)
%learn from selected histograms
%Matlab Svm
% model = svmtrain(sub_treF,sub_treA);
% predicted = svmclassify(model, sub_tesF);
%libsvm
% model = svmtrain(sub_treA,sub_treF, '-s 1 -c 0.5 -n 0.0001');
% predicted = svmpredict(sub_tesA, sub_tesF, model);
%Matlab Boosting
% t = ClassificationTree.template('minleaf',5);
% ens = fitensemble(sub_treF,sub_treA,'AdaBoostM1',1000,t,'LearnRate',0.1,'nprint',100);
% predicted = predict(ens,sub_tesF);