www.pudn.com > gmm_utilities.zip > approximate_gauss_by_gmm.m

function g = approximate_gauss_by_gmm(x,P,N,type) 
%function g = approximate_gauss_by_gmm(x,P,N,type) 
% 
% INPUTS:  
%   x,P - mean and covariance matrix of a Gaussian 
%   N - number of components in output gmm 
%   type - method used 
% 
% OUTPUT: 
%   g - gmm approximation of Gaussian 
% 
% This function is in alpha stage of development. 
% 
% Tim Bailey 2005. 
 
if N == 1, type = 1; elseif nargin == 3, type = 2; end 
 
switch type 
case 1 
    % Trivial solution 
    g.w = ones(1,N)/N; 
    g.x = repcol(x,N); 
    g.P = repmat(P,[1,1,N]); 
case 2 
    % Kernels 
    g = approximate_gauss_by_kernels(x,P,N); 
    g.P = repmat(g.P, [1,1,N]); 
otherwise 
    error('Invalid type') 
end 
     
% Other ideas: 
% - splitting algorithm 
%       - using musso's criterion at each step ?? 
%       - split along principal axis, 1/2 covariance at each step