chmm_def.m

一个隐马尔可夫模型工具箱一个隐马尔可夫模型工具箱一个隐马尔可夫模型工具箱

%---------------------------------------------------------------------------
%
%                           %%%%%%%%%%% 	                    
%                           %CHMM_DEF %
%                           %%%%%%%%%%%	
% 
%
% This function is described only as example to set up the HMM.
% 
% function chmm_def(fhmm)
% 
% This function defines the parameters of the discrete HMM. We can split the parameters in 4 groups:
% 
% Entry:	fhmm is the HMM磗 file name.
% 
% 	vDB defines the database:
% vDB=[' nc ng agrup Np'];
% where nc is the number of class, ng is the number of group,
% agrup defines how to gather the parameters together
% NP is a vector with the number of each group of parameter.
% We have the following relations:
% [nc,ng]=size(vl);
% agrup{ig}=[1 .... size(vl{1,ig}{1},2)+1];
% Np(ig)=length(agrup{ig})-1;
% 
% 	vHMM defines the HMM :
% 	vHMM =[' BAKIS salto maxiter umbral maxitermi Ngauss Ne A B Med Var Pi'];
% if Bakis = 1 it implements a Bakis HMM (also called left-right), else an ergodic HMM is defined 
% Salto is the maximum path authorized in the Bakis HMM from a state to another.
% Maxiter is the maximum iteration number in the Bakis HMM (condition to stop the algorithm)
% Umbral is the threshold condition to stop the HMM (the error is calculated with the maximum likelihood criterion) 
% Maxitermi is the iteration number to find the initial model.
% Ne states number 
% Ngauss is the number of Gaussians in the mixture of Gaussians used to represent the distribution of the observation by state.
% A, B and Pi are the matrices of the HMM (it will be calculated thanks to the CHMM) but must de defined here.
% Med and Var are the matrices of the mean and variances of the Gaussians.
% 
%  vTEST is the matrix parameter for the test :
% vTEST=[' salhmm'];
% Salhmm is a matrix used to store the probabilities values
% We have the following relations:
% salhmm=cell(nc,ng);
%
%
%
% -------------------------------------------------------------------------


function chmm_def(fhmm)

% name of the file of the CHMM
%fhmm='hmm';

iniciar=1;

% VARIABLES of the Database
vDB=[' nc ng agrup Np'];
% classes number and groups number with 
% the relation [nc,ng]=size(vl);
nc=4;
ng=2;

% Definition and numbers of the inputs.
% each realisation is a group of vectors and every vector contains
% a sequence of components that we gather in oreder to create the paramenters inputs.
% We define the gathering of the components and the number of groups.
% we have the relation agrup{ig}=[1 .... size(vl{1,ig}{1},2)+1];
grup=cell(ng,1);
Np=zeros(ng,1);
for ig=1:ng
   agrup{ig}=[1 3];
end
agrup{2}=[1 2 3];
for ig=1:ng
   Np(ig)=length(agrup{ig})-1;
end



% VARIABLES of the HMM
vHMM=[' BAKIS salto maxiter umbral maxitermi Ngauss Ne A B Med Var Pi men'];
% Number of gaussians by groups and parameter.
% All the states have the same number of gaussians
Ngauss=cell(ng,1);
for ig=1:ng   
   Ngauss{ig}=6.*ones(Np(ig),1);% We can change it and fix a number different for each prameter of each group For instance: Ngauss{1}=[1 ; 3];
end;

men=1;
% Variables for the HMM
% States number.
Ne=12.*ones(nc,ng);% we could change the number of state for each class and groupt. Ne=[3 4; 5 6; 7 8; 9 10]
% if Bakis = 1 it implements a Bakis HMM (also called left-right), else an ergodic HMM is defined.
BAKIS=1;
% Salto is the maximum path authorized in the Bakis HMM from a state to another.
salto=1;
% Maximum number of iterations in the Bakis HMM (condition to stop the algorithm).
maxiter=10;
% the threshold condition to stop the HMM (the error is calculated with the maximum likelihood criterion) 
umbral=0.05;
% the iteration number to find the initial model.
maxitermi=5;
% Matrices of the HMM
A=cell(nc,ng);
B=cell(nc,ng);
Med=cell(nc,ng);
Var=cell(nc,ng);
Pi=cell(nc,ng);
for ic=1:nc,
   for ig=1:ng
      A{ic,ig}=zeros(Ne(ic,ig),Ne(ic,ig));
      B{ic,ig}=cell(Np(ig),1);
      Med{ic,ig}=cell(Np(ig),1);
      Var{ic,ig}=cell(Np(ig),1);      
      for ip=1:Np(ig)
         B{ic,ig}{ip}=cell(Ne(ic,ig),1);
         Med{ic,ig}{ip}=cell(Ne(ic,ig),1);
         Var{ic,ig}{ip}=cell(Ne(ic,ig),1);
         for ie=1:Ne(ic,ig),
            B{ic,ig}{ip}{ie}=zeros(Ngauss{ig}(ip),1);      
            Med{ic,ig}{ip}{ie}=zeros(Ngauss{ig}(ip),agrup{ig}(ip+1)-agrup{ig}(ip));
            Var{ic,ig}{ip}{ie}=zeros(Ngauss{ig}(ip),agrup{ig}(ip+1)-agrup{ig}(ip));
         end
      end
      Pi{ic,ig}=zeros(Ne(ic,ig),1);
   end
end


% VARIABLES for the TEST
vTEST=[' salhmm'];

% outputs vectors for each sequence of the test/model
salhmm=cell(nc,ng);

% we save the HMM
guardar=['save ',fhmm,vDB,vHMM,vTEST,' iniciar vDB vHMM vTEST guardar'];
eval([guardar]);