-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathSVM.m
187 lines (160 loc) · 8.35 KB
/
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
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
clc;
clear all;
close all;
tic
fprintf('-----已开始请等待-----\n\n');
%%使用全部10维
% 训练数据
% train_baiyeku=xlsread('.\（new）CCA融合特征参数(标准化数据)\baiyekuTrain.xlsx',1,'C2:L2001');
% train_daowenbing=xlsread('.\（new）CCA融合特征参数(标准化数据)\daowenbingTrain.xlsx',1,'C2:L2001');
% train_humaban=xlsread('.\（new）CCA融合特征参数(标准化数据)\humabanTrain.xlsx',1,'C2:L2001');
% % 测试数据
% test_baiyeku=xlsread('.\（new）CCA融合特征参数(标准化数据)\baiyekuTest.xlsx',1,'C2:L501');
% test_baiyeku_label=xlsread('.\（new）CCA融合特征参数(标准化数据)\baiyekuTest.xlsx',1,'B2:B501');
% test_daowenbing=xlsread('.\（new）CCA融合特征参数(标准化数据)\daowenbingTest.xlsx',1,'C2:L501');
% test_daowenbing_label=xlsread('.\（new）CCA融合特征参数(标准化数据)\daowenbingTest.xlsx',1,'B2:B501');
% test_humaban=xlsread('.\（new）CCA融合特征参数(标准化数据)\humabanTest.xlsx',1,'C2:L501');
% test_humaban_label=xlsread('.\（new）CCA融合特征参数(标准化数据)\humabanTest.xlsx',1,'B2:B501');
%
% %使用6维
% %训练数据
% train_baiyeku=xlsread('.\（new）CCA融合特征参数(标准化数据)\baiyekuTrain.xlsx',1,'C2:H2001');
% train_daowenbing=xlsread('.\（new）CCA融合特征参数(标准化数据)\daowenbingTrain.xlsx',1,'C2:H2001');
% train_humaban=xlsread('.\（new）CCA融合特征参数(标准化数据)\humabanTrain.xlsx',1,'C2:H2001');
%
% % 测试数据
% test_baiyeku=xlsread('.\（new）CCA融合特征参数(标准化数据)\baiyekuTest.xlsx',1,'C2:H501');
% test_baiyeku_label=xlsread('.\（new）CCA融合特征参数(标准化数据)\baiyekuTest.xlsx',1,'B2:B501');
% test_daowenbing=xlsread('.\（new）CCA融合特征参数(标准化数据)\daowenbingTest.xlsx',1,'C2:H501');
% test_daowenbing_label=xlsread('.\（new）CCA融合特征参数(标准化数据)\daowenbingTest.xlsx',1,'B2:B501');
% test_humaban=xlsread('.\（new）CCA融合特征参数(标准化数据)\humabanTest.xlsx',1,'C2:H501');
% test_humaban_label=xlsread('.\（new）CCA融合特征参数(标准化数据)\humabanTest.xlsx',1,'B2:B501');
% %使用直接融合
%训练数据
train_baiyeku=xlsread('4CCA融合特征(原始数据).xlsx',1,'C1:D2000');
train_daowenbing=xlsread('4CCA融合特征(原始数据).xlsx',1,'C2501:D4500');
train_humaban=xlsread('4CCA融合特征(原始数据).xlsx',1,'C5001:D7000');
% 测试数据
test_baiyeku=xlsread('4CCA融合特征(原始数据).xlsx',1,'C2001:D2500');
test_baiyeku_label=xlsread('4CCA融合特征(原始数据).xlsx',1,'B2001:B2500');
test_daowenbing=xlsread('4CCA融合特征(原始数据).xlsx',1,'C4501:D5000');
test_daowenbing_label=xlsread('4CCA融合特征(原始数据).xlsx',1,'B4501:B5000');
test_humaban=xlsread('4CCA融合特征(原始数据).xlsx',1,'C7001:D7500');
test_humaban_label=xlsread('4CCA融合特征(原始数据).xlsx',1,'B7001:B7500');
%使用原始图片直接融合
%训练数据
% train_baiyeku=xlsread('newOriginalFeature.xlsx',1,'C1:AT11');
% train_daowenbing=xlsread('newOriginalFeature.xlsx',1,'C12:AT21');
% train_humaban=xlsread('newOriginalFeature.xlsx',1,'C22:AT30');
%
% % 测试数据
% test_baiyeku=xlsread('newOriginalFeature.xlsx',1,'C1:AT11');
% test_baiyeku_label=xlsread('newOriginalFeature.xlsx',1,'B1:B11');
% test_daowenbing=xlsread('newOriginalFeature.xlsx',1,'C12:AT21');
% test_daowenbing_label=xlsread('newOriginalFeature.xlsx',1,'B12:B21');
% test_humaban=xlsread('newOriginalFeature.xlsx',1,'C22:AT30');
% test_humaban_label=xlsread('newOriginalFeature.xlsx',1,'B22:B30');
% 测试数据汇总
test_features=[test_baiyeku;test_daowenbing;test_humaban];
% 测试数据真实标签汇总
test_labels = [test_baiyeku_label;test_daowenbing_label;test_humaban_label];
%%
% 训练数据分为3类
% 类别i的 正样本 选择类别i的全部，负样本 从其余类别中随机选择（个数为正样本相同）
% 类别1:白叶枯
baiyeku_p = train_baiyeku;
number_p=size(train_baiyeku,1);
% 其余类别训练样本汇总
others=[train_daowenbing;train_humaban];
number_o=size(others,1);
% randperm(n,k)是从1到n的序号中随机返回k个
index1 = randperm(number_o,number_p);
% 从其余样本中随机选择k个
baiyeku_n = others(index1,:);
train_bai = [baiyeku_p;baiyeku_n];
% 正类表示为1，负类表示为-1
train_bai_label = [ones(number_p,1);-1*ones(number_p,1)];
% 类别2:稻瘟病
daowenbing_p = train_daowenbing;
number_p=size(train_daowenbing,1);
% 其余类别训练样本汇总
others=[train_baiyeku;train_humaban];
number_o=size(others,1);
% randperm(n,k)是从1到n的序号中随机返回k个
index1 = randperm(number_o,number_p);
% 从其余样本中随机选择k个
daowenbing_n = others(index1,:);
train_dao = [daowenbing_p;daowenbing_n];
% 正类表示为1，负类表示为-1
train_dao_label = [ones(number_p,1);-1*ones(number_p,1)];
% 类别3:胡麻斑
humaban_p = train_humaban;
number_p=size(train_humaban,1);
% 其余类别训练样本汇总
others=[train_baiyeku;train_daowenbing];
number_o=size(others,1);
% randperm(n,k)是从1到n的序号中随机返回k个
index1 = randperm(number_o,number_p);
% 从其余样本中随机选择k个
humaban_n = others(index1,:);
train_hu = [humaban_p;humaban_n];
% 正类表示为1，负类表示为-1
train_hu_label = [ones(number_p,1);-1*ones(number_p,1)];
%
%分别训练3个类别的SVM模型
%OptimizeHyperparameters超参数优化
% model_baiyeku = fitcsvm(train_bai,train_bai_label,'ClassNames',{'-1','1'},'Standardize',true,'KernelFunction','rbf','OptimizeHyperparameters',{'BoxConstraint','KernelScale'},'HyperparameterOptimizationOptions',struct('ShowPlots',false) );
% model_daowenbing = fitcsvm(train_dao,train_dao_label,'ClassNames',{'-1','1'},'Standardize',true,'KernelFunction','rbf','OptimizeHyperparameters',{'BoxConstraint','KernelScale'},'HyperparameterOptimizationOptions',struct('ShowPlots',false));
% model_humaban = fitcsvm(train_hu,train_hu_label,'ClassNames',{'-1','1'},'Standardize',true,'KernelFunction','rbf','OptimizeHyperparameters',{'BoxConstraint','KernelScale'} ,'HyperparameterOptimizationOptions',struct('ShowPlots',false));
% fprintf('-----模型训练完毕-----\n\n');
% % % % 带入训练好的三个模型的BoxConstraint和KernelScale两个参数并运行
model_baiyeku = fitcsvm(train_bai,train_bai_label,'ClassNames',{'-1','1'},'Standardize',true,'KernelFunction','rbf','BoxConstraint', 3.71826944713741,'KernelScale', 0.473812982562996);
model_daowenbing = fitcsvm(train_dao,train_dao_label,'ClassNames',{'-1','1'},'Standardize',true,'KernelFunction','rbf','BoxConstraint', 12.7016070067197 ,'KernelScale',0.360670302805846);
model_humaban = fitcsvm(train_hu,train_hu_label,'ClassNames',{'-1','1'},'Standardize',true,'KernelFunction','rbf','BoxConstraint',8.81864834274725 ,'KernelScale',0.260815633743261);
fprintf('-----模型训练完毕-----\n\n');
%%
% 分别保存3个SVM模型
% saveCompactModel(model_baiyeku,'SVM_baiyeku5.mat');
% saveCompactModel(model_daowenbing,'SVM_daowenbing5.mat');
% saveCompactModel(model_humaban,'SVM_humaban5.mat');
save('SVM_baiyeku7.mat', 'model_baiyeku');
save('SVM_daowenbing7.mat', 'model_daowenbing');
save('SVM_humaban7.mat', 'model_humaban');
fprintf('-----模型保存完毕-----\n\n');
%%
% label是n*1的矩阵，每一行是对应测试样本的预测标签；
% score是n*2的矩阵，第一列为预测为“负”的得分，第二列为预测为“正”的得分。
% 用训练好的3个SVM模型分别对测试样本进行预测分类，得到3个预测标签
[predict_baiyeku,score_baiyeku] = predict(model_baiyeku,test_features);
[predict_daowenbing,score_daowenbing] = predict(model_daowenbing,test_features);
[predict_humaban,score_humaban] = predict(model_humaban,test_features);
% 求出测试样本在3个模型中预测为“正”得分的最大值，作为该测试样本的最终预测标签
score = [score_baiyeku(:,2),score_daowenbing(:,2),score_humaban(:,2)];
% 最终预测标签
final_labels = zeros(size(test_features,1),1);
for i = 1:size(final_labels,1)
% 返回每一行的最大值和其位置
[m,p] = max(score(i,:));
% 位置即为标签
final_labels(i,:) = p;
end
fprintf('-----样本预测完毕-----\n\n');
% 分类评价指标
group = test_labels; % 真实标签
grouphat = final_labels; % 预测标签
% C为混淆矩阵，C(i,j)表示在g1中的i类数在g2中被分在j类的个数，即真实标签被分到的测试标签
[C,order] = confusionmat(group,grouphat,'Order',[1;2;3]); % 'Order'指定类别的顺序
c1_p = C(1,1) / sum(C(:,1));
c1_r = C(1,1) / sum(C(1,:));
c1_F = 2*c1_p*c1_r / (c1_p + c1_r);
fprintf('白叶枯类的查准率为%f,查全率为%f,F测度为%f\n\n',c1_p,c1_r,c1_F);
c2_p = C(2,2) / sum(C(:,2));
c2_r = C(2,2) / sum(C(2,:));
c2_F = 2*c2_p*c2_r / (c2_p + c2_r);
fprintf('稻瘟病类的查准率为%f,查全率为%f,F测度为%f\n\n',c2_p,c2_r,c2_F);
c3_p = C(3,3) / sum(C(:,3));
c3_r = C(3,3) / sum(C(3,:));
c3_F = 2*c3_p*c3_r / (c3_p + c3_r);
fprintf('胡麻斑类的查准率为%f,查全率为%f,F测度为%f\n\n',c3_p,c3_r,c3_F);
c_p=(C(1,1)+C(2,2)+C(3,3))/size(test_features,1);
fprintf('总识别准确率为%f\n',c_p);