Download All

Code covered by the BSD License  

Highlights from
Free Split and Merge Expectation Maximization for MultiVariate Gaussian Mixture

image thumbnail
from Free Split and Merge Expectation Maximization for MultiVariate Gaussian Mixture by Sebastien PARIS
FSMEM can estimate MVGM parameters and number of conpounds via split/merge cluster moves

gene_mvgm(d , K , sig) 
function [M , S , P] = gene_mvgm(d , K , sig)

%
%
% [M , S , P] = gene_mvgm(d , K , sig);
%
%
%


eyed   = eye(d);


S      = zeros(d , d , K);


P      = rand(1 , 1 , K);
sP     = sum(P , 3);
P      = P./sP(: , : , ones(1  , K));

M      = 1.25*randn(d , 1 , K);


for k=1:K
      
    % Compute variance matrix
    
    Rot = eyed;
    for dd=1:d
        for e=(dd+1):d
            theta      = rand(1) * 2*pi;
            rot        = eyed;
            rot(dd,dd) = cos(theta);
            rot(e,e)   = rot(d,d);
            rot(dd,e)  = sin(theta);
            rot(e,dd)  = -rot(d,e);
            Rot        = rot*Rot;
        end
    end

    Rot          = sig*Rot*diag(randn(d,1) + 1);
    S(: , : , k) = Rot*Rot';
   
end

Contact us at files@mathworks.com