dpOverMap

DP over matrix of state probability.

Contents

Syntax

Description

[optimValue, dpPath]=dpOverMap(stateProbMat, transProbMat) returns the optimum value and the corresponding DP (dynamic programming) for HMM evaluation.

• stateProbMat: matrix of log state probabilities
• transProbMat: matrix of log transition probabilities

Example

load pfMat.mat
pfMat(1:20, :)=0;
%pfMat=[5 2 6; 2 9 3];
[stateNum, frameNum]=size(pfMat);
penalty=10000;
opt.initProb=log(1/stateNum)*ones(1, stateNum);
%opt.initProb=zeros(1, stateNum); opt.initProb(1)=1; opt.initProb=log(opt.initProb);
opt.endState=logical(ones(1, stateNum));
%opt.endState=logical(zeros(1, stateNum)); opt.endState(end)=1;
showPlot=1;
[optimValue, dpPath, dpTable, time]=dpOverMap(pfMat, penalty, opt, showPlot);
fprintf('Time=%.2f sec\n', time);

Time=0.02 sec

See Also

dpOverMapM.

Published with MATLAB® R2018a