-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathrun_demo.m
257 lines (232 loc) · 8.34 KB
/
run_demo.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
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
clear;
clc;
stream = RandStream.getGlobalStream;
reset(stream);
addpath(genpath('ClusteringMeasure'));
addpath(genpath('.\SS-US-ELM'));
addpath(genpath('.\RMSC-ELM'));
addpath(genpath('.\MMSC-ELM'));
addpath(genpath('.\MKC-LKAM'));
addpath(genpath('SSC_1.0'));
addpath(genpath('.\DisClusteringCode'));
addpath(genpath('C:\Program Files\Mosek\7\toolbox\r2013a'));
datapath = 'F:\博士阶段\多视图ELM后期融合\无监督聚类\dataset\singleview\';
load([datapath,'iris'],'data','label');
%load([datapath,'heart'],'data','label');
%load([datapath,'libras'],'data','label');
%load([datapath,'balance'],'data','label');
%load([datapath,'diabetes'],'data','label');
%load([datapath,'vehicle'],'data','label');
%load([datapath,'CNAE'],'data','label');
%load([datapath,'segment'],'data','label');
%load([datapath,'yale'],'data','label');
%load([datapath,'orl'],'data','label');
x = transpose(mapminmax(data',-1,1));
num_class = length(unique(label));
%% Kmeans
fprintf('Result of Kmeans\n');
[F,P,R,nmi,avgent,AR,ACC,Purity]=MyKmeans(x,num_class,label);
fprintf('======================================\n');
%% Kernel kmeans without ELM
% fprintf('Result of KKM without ELM\n');
% tic;
% options.KernelType = 'Gaussian';
% % options.t=optSigma(x);%sigma取平均值
% lambda_values = [];
% lambda_ii = [];
% ACC_avg = [];
% ACC_std = [];
% nmi_avg = [];
% nmi_std = [];
% Purity_avg = [];
% Purity_std = [];
%
% for ii=-10:10
% lambda_values = [lambda_values,2^ii];
% lambda_ii = [lambda_ii,ii];
% end
% for i=1:length(lambda_values)
% options.t=lambda_values(i);
% myK = constructKernel(x,x,options);
% K = (myK+myK')/2 + 1e-8*eye(size(myK));
% opt.disp = 0;
% [H,~] = eigs(K,num_class,'la',opt);
% H_normalized = H;
% [res_kkm] = accuFucV2(H_normalized,label,num_class);
%
% ACC_avg =[ACC_avg, res_kkm.Acc_avg];
% ACC_std = [ACC_std,res_kkm.Acc_std];
% nmi_avg = [nmi_avg,res_kkm.NMI_avg];
% nmi_std = [nmi_std,res_kkm.NMI_std];
% Purity_avg =[Purity_avg, res_kkm.Purity_avg];
% Purity_std =[Purity_std, res_kkm.Purity_std];
% end
% disp(['SSC运行时间:',num2str(toc),'s']);
% step_ACC = lambda_ii(find(ACC_avg ==max(ACC_avg)));
% step_NMI = lambda_ii(find(nmi_avg ==max(nmi_avg)));
% step_Purity =lambda_ii(find(Purity_avg ==max(Purity_avg)));
% fprintf('step_ACC=%f, step_NMI=%f, step_Purity=%f\n',step_ACC(1),step_NMI(1),step_Purity(1));
% max_acc_std = ACC_std(find(ACC_avg==max(ACC_avg)));
% max_nmi_std = nmi_std(find(nmi_avg==max(nmi_avg)));
% max_Purity_std = Purity_std(find(Purity_avg==max(Purity_avg)));
% fprintf('ACC=%0.4f(%0.4f),nmi score=%0.4f(%0.4f),Purity=%0.4f(%0.4f)\n',max(ACC_avg),max_acc_std(1),max(nmi_avg),max_nmi_std(1),max(Purity_avg),max_Purity_std(1));
%
% fprintf('======================================\n');
%% Spectral clustering on kernel without ELM
% fprintf('Result of Spectral clusteirng on kernel withou ELM\n')
% tic;
% options.KernelType = 'Gaussian';
% % options.t=optSigma(x);
% lambda_values = [];
% lambda_ii = [];
% ACC_avg = [];
% ACC_std = [];
% nmi_avg = [];
% nmi_std = [];
% Purity_avg = [];
% Purity_std = [];
%
% for ii=-10:10
% lambda_values = [lambda_values,2^ii];
% lambda_ii = [lambda_ii,ii];
% end
% for i=1:length(lambda_values)
% options.t=lambda_values(i);
% K_sok = constructKernel(x,x,options);
% [~,~,~,~,~,nmi_sok,~,~,~,ACC_sok,Purity_sok] = baseline_spectral_onkernel(K_sok,num_class,label);
%
% ACC_avg =[ACC_avg, ACC_sok(1)];
% ACC_std = [ACC_std,ACC_sok(2)];
% nmi_avg = [nmi_avg,nmi_sok(1)];
% nmi_std = [nmi_std,nmi_sok(2)];
% Purity_avg =[Purity_avg, Purity_sok(1)];
% Purity_std =[Purity_std, Purity_sok(2)];
% end
% disp(['SSC运行时间:',num2str(toc),'s']);
% step_ACC = lambda_ii(find(ACC_avg ==max(ACC_avg)));
% step_NMI = lambda_ii(find(nmi_avg ==max(nmi_avg)));
% step_Purity =lambda_ii(find(Purity_avg ==max(Purity_avg)));
% fprintf('step_ACC=%f, step_NMI=%f, step_Purity=%f\n',step_ACC(1),step_NMI(1),step_Purity(1));
% max_acc_std = ACC_std(find(ACC_avg==max(ACC_avg)));
% max_nmi_std = nmi_std(find(nmi_avg==max(nmi_avg)));
% max_Purity_std = Purity_std(find(Purity_avg==max(Purity_avg)));
% fprintf('ACC=%0.4f(%0.4f),nmi score=%0.4f(%0.4f),Purity=%0.4f(%0.4f)\n',max(ACC_avg),max_acc_std(1),max(nmi_avg),max_nmi_std(1),max(Purity_avg),max_Purity_std(1));
%
% disp(['SCoK运行时间:',num2str(toc),'s']);
% fprintf('======================================\n');
%% SSC
% fprintf('Result of SSC without ELM\n');
% tic;
% X = x';
% r = 0; %Enter the projection dimension e.g. r = d*n, enter r = 0 to not project
% Cst = 0; %Enter 1 to use the additional affine constraint sum(c) == 1
% OptM = 'Lasso'; %OptM can be {'L1Perfect','L1Noise','Lasso','L1ED'}
% lambda = 0.01; %Regularization parameter in 'Lasso' or the noise level for 'L1Noise'
% K = 0;%max(d1,d2); %Number of top coefficients to build the similarity graph, enter K=0 for using the whole coefficients
% Xp = DataProjection(X,r,'NormalProj');
% CMat = SparseCoefRecovery(Xp,Cst,OptM,lambda);
%
% options.KernelType = 'Gaussian';
% options.t=optSigma(CMat);
% % options.t=2^(9);
% K_ssc = constructKernel(CMat,CMat,options);
% [~,~,~,~,~,nmi_ssc,~,~,~,ACC_ssc,Purity_ssc] = baseline_spectral_onkernel(K_ssc,num_class,label);
% disp(['SSC运行时间:',num2str(toc),'s']);
% fprintf('======================================\n');
%% ELM kmeans
% fprintf('Result of ELM Kmenas\n');
% L = 2000;
% [H,OutputWeight]=myelm(x,label,L, 'sigmoid');
% fprintf('Nodes of hidden layers:%d\n',L);
% [F,P,R,nmi,avgent,AR,ACC,Purity]=MyKmeans(H,num_class,label);
% fprintf('======================================\n');
%% US-ELM
% fprintf('Result of US-ELM\n');
% hidden =2000;
% MyUSELM_kmeans(x,num_class,label,hidden);
% fprintf('======================================\n');
%% Iter welm
% fprintf('Result of Iter WELM\n');
% hidden=2000;
% MyIterWELM(x,num_class,label,hidden);
% fprintf('======================================\n');
%% ELMC LDA
%letterAB
% fprintf('Result of ELMC LDA\n');
% hidden=2000;
% My_ELMC_LDA(x,num_class,label,hidden);
% fprintf('======================================\n');
%% ELMC KM
%yaleb
% fprintf('Result of ELMC KM\n');
% hidden =2000;
% My_ELMC_KM(x,num_class,label,hidden);
% fprintf('======================================\n');
%% ELM baseline
% fprintf('Result of ELM baseline\n');
% num_views=20;
% sigma_value=2^(10);
% MyELM_baseline(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% ELM Co-Reg Spectral
%yaleb
% fprintf('Result of ELM Co-Reg Spectral\n');
% num_views =20;
% sigma_value=2^(8);
% ELM_CoReg_result = ELM_Co_Reg_spectral(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% ELM RMSC
%yaleb
% fprintf('Result of ELM RMSC\n');
% num_views=20;
% sigma_value=2^(10);
% ELM_RMSC_result = MyELM_RMSC(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% ELM MMSC
% fprintf('Result of ELM MMSC\n');
% num_views =20;
% MyELM_MMSC(x,num_class,label,num_views);
% fprintf('======================================\n');
%% ELMC MKKM-MR
% fprintf('Result of MMKM-MR\n');
% num_views = 20;
% sigma_value=2^(0);
% MyELM_MMKM_MR(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% ELMC MKKM-LKAM
%yale 2^0 1 2 3 4已测
% fprintf('Result of MMV LKAM\n');
% num_views = 20;
% sigma_value=2^(5);
% MyELM_MKC_LKAM(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% A-MKKM
% fprintf('Result of A-MKKM\n');
% num_views = 20;
% sigma_value=2^(4);
% mykernelkmeans(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% SB-KKM
% fprintf('Result of SB-KKM\n');
% num_views = 20;
% sigma_value=2^(2);
% mySBkernelkmeans(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% MKKM
% fprintf('Result of MKKM\n');
% num_views = 20;
% sigma_value=2^(4);
% mkkmeans_train(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% LMKKM
% fprintf('Result of LMKKM\n');
% num_views = 20;
% sigma_value=2^(3);
% lmkkmeans_train(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');
%% RMKKM
% fprintf('Result of LMKKM\n');
% num_views = 20;
% sigma_value=2^(8);
% my_RMKKM(x,num_class,label,num_views,sigma_value);
% fprintf('======================================\n');