-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathSVM_3_CLASSIFY.m
189 lines (149 loc) · 6 KB
/
SVM_3_CLASSIFY.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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
%% classify using Support Vector Machine
%classifier1：Avian(0)——Swine(2）
%classifier2: Avian(0)——Human(1）
%classifier3：Swine(2)——Human(1）
load AvianHumanSwine_HA_influenzafaa_2_sample;
outputfile = 'AvianHumanSwine_HA_svm_influenzafaa_data_2_10foldresult.xls';
%%
ACCURACY = zeros(1,10);
%
avian_respective_accuracy = zeros(1,10);
human_respective_accuracy = zeros(1,10);
swine_respective_accuracy = zeros(1,10);
%%
Avian_ind = find(cell2mat(sampleIVT(:,102)) == 0);
Avian = sampleIVT(Avian_ind,:);
[Avian_num,C0] = size(Avian);
indicesA = crossvalind('Kfold',Avian_num,10);
Human_ind = find(cell2mat(sampleIVT(:,102)) == 1);
Human = sampleIVT(Human_ind,:);
[Human_num,C1] = size(Human);
indicesH = crossvalind('Kfold',Human_num,10);
Swine_ind = find(cell2mat(sampleIVT(:,102)) == 2);
Swine = sampleIVT(Swine_ind,:);
[Swine_num,C2] = size(Swine);
indicesS = crossvalind('Kfold',Swine_num,10);
%% do ten times
for k=1:10
% cross validation
testA = (indicesA == k);
trainA = ~testA;
Avian_test = Avian(testA,:);
Avian_train = Avian(trainA,:);
testH = (indicesH == k);
trainH = ~testH;
Human_test = Human(testH,:);
Human_train = Human(trainH,:);
testS = (indicesS == k);
trainS = ~testS;
Swine_test = Swine(testS,:);
Swine_train = Swine(trainS,:);
% test sets
TEST = cat(1,Avian_test,Human_test,Swine_test);
[testnum,C3] = size(TEST);
X_test = cell2mat(TEST(:,2:end-1));
Y_test = cell2mat(TEST(:,end));
Y_PREDICT = zeros(testnum,3);
Y_predict = zeros(testnum,1);
% train sets
TRAIN02 = cat(1,Avian_train,Swine_train);
X_train02 = cell2mat(TRAIN02(:,2:end-1));
Y_train02 = cell2mat(TRAIN02(:,end));
TRAIN01 = cat(1,Avian_train,Human_train);
X_train01 = cell2mat(TRAIN01(:,2:end-1));
Y_train01 = cell2mat(TRAIN01(:,end));
TRAIN21 = cat(1,Swine_train,Human_train);
X_train21 = cell2mat(TRAIN21(:,2:end-1));
Y_train21 = cell2mat(TRAIN21(:,end));
% Construct three classifiers of SVM
option_new02 = statset('MaxIter',20000);
SVMStruct02 = svmtrain(X_train02,Y_train02,'kernel_function','rbf','options',option_new02);
Y_PREDICT(:,1) = svmclassify(SVMStruct02,X_test);
option_new01 = statset('MaxIter',20000);
SVMStruct01 = svmtrain(X_train01,Y_train01,'kernel_function','rbf','options',option_new01);
Y_PREDICT(:,2) = svmclassify(SVMStruct01,X_test);
option_new21 = statset('MaxIter',20000);
SVMStruct21 = svmtrain(X_train21,Y_train21,'kernel_function','rbf','options',option_new21);
Y_PREDICT(:,3) = svmclassify(SVMStruct21,X_test);
%% get the label of classify using vote
class = [0,1,2];
for i = 1:testnum
count0 = 0;
count1 = 0;
count2 = 0;
for j = 1:3
if Y_PREDICT(i,j) == 0
count0 = count0+1;
end
if Y_PREDICT(i,j) == 1
count1 = count1+1;
end
if Y_PREDICT(i,j) == 2
count2 = count2+1;
end
end
[count,clsss_ind] = max([count0,count1,count2]);
if count == 1
sjs = randperm(3);
Y_predict(i,1) = class(sjs(1));
else
Y_predict(i,1) = clsss_ind-1;
end
end
%% calculate accuracy rate
Ytp = cat(2,Y_test,Y_predict);
a0_ind = find(Ytp(:,1) == 0);
h1_ind = find(Ytp(:,1) == 1);
s2_ind = find(Ytp(:,1) == 2);
Ya = Ytp(a0_ind,:);
[a0_row,a0_col] = size(Ya);
a0_count = 0;
for ai = 1:a0_row
if Ya(ai,1) == Ya(ai,2)
a0_count = a0_count + 1;
end
end
avian_respective_accuracy(1,k) = a0_count/a0_row;
Yh = Ytp(h1_ind,:);
[h1_row,h1_col] = size(Yh);
h1_count = 0;
for hi = 1:h1_row
if Yh(hi,1) == Yh(hi,2)
h1_count = h1_count + 1;
end
end
human_respective_accuracy(1,k) = h1_count/h1_row;
Ys = Ytp(s2_ind,:);
[s2_row,s2_col] = size(Ys);
s2_count = 0;
for si = 1:s2_row
if Ys(si,1) == Ys(si,2)
s2_count = s2_count + 1;
end
end
swine_respective_accuracy(1,k) = s2_count/s2_row;
% overall accuracy rate
ACCURACY(1,k) = (a0_count + h1_count + s2_count)/(a0_row + h1_row + s2_row);
end
% calculate average value and standard deviation of accuracy rate
accuracy_average = mean(ACCURACY);
accuracy_sd = std(ACCURACY);
%
avian_respective_accuracy_average = mean(avian_respective_accuracy);
avian_respective_accuracy_sd = std(avian_respective_accuracy);
human_respective_accuracy_average = mean(human_respective_accuracy);
human_respective_accuracy_sd = std(human_respective_accuracy);
swine_respective_accuracy_average = mean(swine_respective_accuracy);
swine_respective_accuracy_sd = std(swine_respective_accuracy);
xlswrite(outputfile,ACCURACY,1,'B6');
xlswrite(outputfile,avian_respective_accuracy,1,'B7');
xlswrite(outputfile,human_respective_accuracy,1,'B8');
xlswrite(outputfile,swine_respective_accuracy,1,'B9');
xlswrite(outputfile,accuracy_average,1,'B13');
xlswrite(outputfile,accuracy_sd,1,'C13');
xlswrite(outputfile,avian_respective_accuracy_average,1,'B14');
xlswrite(outputfile,avian_respective_accuracy_sd,1,'C14');
xlswrite(outputfile,human_respective_accuracy_average,1,'B15');
xlswrite(outputfile,human_respective_accuracy_sd,1,'C15');
xlswrite(outputfile,swine_respective_accuracy_average,1,'B16');
xlswrite(outputfile,swine_respective_accuracy_sd,1,'C16');