obslike.m

VARHMMBOX, version 1.1, Iead Rezek, Oxford University, MAR 2002 Matlab toolbox for Hidden Markov Mo

function [B] = obslike (X,T,n,hmm)

% function [B] = obslike (X,T,n,hmm)
%
% Evaluate likelihood of data given observation model
% for hmm.obsmodel = 'GaussCom','Gauss','LIKE'
% 
% X          N by p data matrix
% T          length of series to learn
% n          block index (time series data can be split into many blocks)
% hmm        hmm data structure
%
% B          Likelihood of N data points

p=length(X(1,:));
N=length(X(:,1));
K=hmm.K;

B=zeros(T,K);
k1=(2*pi)^(-p/2);
switch hmm.obsmodel
 case 'GaussCom',  
  iCov=inv(hmm.state(1).Cov);  % All Covs are the same
  k2=k1/sqrt(det(hmm.state(1).Cov));
  for i=1:T
    for l=1:K
      d=hmm.state(l).Mu-X((n-1)*T+i,:);
      B(i,l)=k2*exp(-0.5*d*iCov*d');
    end
  end
 case 'Gauss',
  for l=1:K  
    state(l).iCov=inv(hmm.state(l).Cov);
    state(l).k2=k1/sqrt(det(hmm.state(l).Cov));
  end
  for i=1:T
    for l=1:K
      d=hmm.state(l).Mu-X((n-1)*T+i,:);
      B(i,l)=state(l).k2*exp(-0.5*d*state(l).iCov*d');
    end
  end
 case 'Poisson',
  for l=1:K
    lambda=hmm.state(l).lambda;
    B(:,l)=lambda.^X(:,2).*exp(-X(:,1).*lambda)./gamma(X(:,2)+1);
  end
 case 'LIKE',
  % The observations are themselves likelihoods
  for l=1:K
    B(:,l)=X(:,l);
  end    
 otherwise
  disp('Unknown observation model');
end