TTB_modellingvortexinteraction/Analysis_Ceff.m at main · decolab/TTB_modellingvortexinteraction · GitHub

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
clear all;
load model_vortex_partition_rest_vs_task.mat;

NSUB=989;
NP=100;

gbc_ceff_rest_sub=mean(Ceff_rest_sub,3);
gbc_ceff_social_sub=mean(Ceff_social_sub,3);

nn=1;
for i=1:NP
    a=gbc_ceff_rest_sub(:,i)';
    b=gbc_ceff_social_sub(:,i)';
    pp(nn)=ranksum(a,b);
    nn=nn+1;
end
nsig=FDR_benjHoch(pp,0.05);


%%%  Classification

NTRAIN=900;
kfold=1000;
cl=1:2;
pc=zeros(2,2);
for nfold=1:kfold
    shuffling=randperm(NSUB);
    TrainData1=[];
    for sub=shuffling(1:NTRAIN)
        TS=gbc_ceff_rest_sub(sub,nsig);
        TrainData1=vertcat(TrainData1,TS);
    end
    XValidation1=[];
    for sub=shuffling(NTRAIN+1:end)
        TS=gbc_ceff_rest_sub(sub,nsig);
        XValidation1=vertcat(XValidation1,TS);
    end
    Responses1=categorical(ones(size(TrainData1,1),1),cl);
    YValidation1=categorical(ones(size(XValidation1,1),1),cl);

    TrainData2=[];
    for sub=shuffling(1:NTRAIN)
        TS=gbc_ceff_social_sub(sub,nsig);
        TrainData2=vertcat(TrainData2,TS);
    end
    XValidation2=[];
    for sub=shuffling(NTRAIN+1:end)
        TS=gbc_ceff_social_sub(sub,nsig);
        XValidation2=vertcat(XValidation2,TS);
    end
    Responses2=categorical(2*ones(size(TrainData2,1),1),cl);
    YValidation2=categorical(2*ones(size(XValidation2,1),1),cl);

    TrainData=vertcat(TrainData1,TrainData2);
    XValidation=vertcat(XValidation1,XValidation2);
    Responses=vertcat(Responses1,Responses2);
    YValidation=vertcat(YValidation1,YValidation2);

    t = templateSVM('KernelFunction','rbf');
    svmmodel=fitcecoc(TrainData,Responses,'Learners',t);

    % compute
    valno1=size(XValidation1,1);
    con=zeros(2,2);
    test1=predict(svmmodel,XValidation1);
    for i=1:valno1
        winclass=test1(i);
        con(1,winclass)=con(1,winclass)+1;
    end
    valno2=size(XValidation2,1);
    test2=predict(svmmodel,XValidation2);
    for i=1:valno2
        winclass=test2(i);
        con(2,winclass)=con(2,winclass)+1;
    end
    con(1,:)=con(1,:)/valno1;
    con(2,:)=con(2,:)/valno2;
    accdist(nfold)=sum(diag(con))/2;
    pc=pc+con;
end
pc=pc/kfold
acc=sum(diag(pc))/2