cvLabMATLAB/orzKPCA.m at main · ComputerVisionLaboratory/cvLabMATLAB · 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
function [A,D,C,K] = orzKPCA(X,Y,nSigma,varargin)


X = X(:,:);
[nDim,nNum] =size(X);

flgM = true;
useRFF = false;
numRandFeats = 0;
if nargin == 4
    if varargin{1} == 'R'
        flgM = false;
    end
end
if nargin > 4
    if varargin{1} == 'R'
        flgM = false;
    end
    if varargin{2} > 0
        useRFF = true;
        numRandFeats = varargin{2};
    end
end

% K = zeros(nNum,nNum);

if useRFF
    K = cvtKernelGramMatrixRFF(X, numRandFeats, nSigma);
else
    K = cvtKernelGramMatrix(X, nSigma);
end


if flgM==true
    IN =  ones(nNum,nNum)/nNum;
    K = K - IN*K - K*IN + IN*K*IN;
end


if Y <= 1 % ��^��
    cRate = Y;
    [A B] = eig(K);
    [B ind] = sort(diag(B),'descend');
    A=A(:,ind);
    D = B/nNum;
    nSubDim = find(cumsum(D)/sum(D)>=cRate, 1 );
    A=A(:,1:nSubDim);
    B=B(1:nSubDim);
    A = A/sqrt(diag(B));
    %U = X*A;
    D = D(1:nSubDim);
    C = sum(D);

elseif Y > 1 % �听����Ԃ̎���
    nSubDim = floor(Y);
    OPTS.disp = 0;
    [A B] = eigs(K,nSubDim,'lm',OPTS);  %[A B] = eigs(K,nSubDim);
    [B ind] = sort(diag(B),'descend');
    A=A(:,ind);
    D = B/nNum;
    A = A/sqrt(diag(B));
    C = sum(D);
end