myPlateauPass
An m-file implementation of DP over matrix of state probability.
Contents
Syntax
- optimValue=dpOverMapM(P, transProbMat)
- optimValue=dpOverMapM(P, transProbMat, showPlot)
- [optimValue, dpPath, dpTable, time]=dpOverMapM(...)
Description
[optimValue, dpPath]=dpOverMap(P, transProbMat) returns the optimum value and the corresponding DP (dynamic programming) for HMM evaluation.
- P: matrix of log state probabilities
- transProbMat: matrix of log transition probabilities
Example
m=320;
n=120;
P=rand(m, n);
[stateNum, frameNum]=size(P);
alpha=0;
showPlot=1;
[optimValue, dpPath, dpTable, time]=myPlateauPass(P, alpha, showPlot);
fprintf('Time=%.2f sec\n', time);
Time=0.08 sec
