Mercurial > hg > camir-aes2014
annotate toolboxes/FullBNT-1.0.7/HMM/mdp_sample.m @ 0:e9a9cd732c1e tip
first hg version after svn
| author | wolffd |
|---|---|
| date | Tue, 10 Feb 2015 15:05:51 +0000 |
| parents | |
| children |
| rev | line source |
|---|---|
| wolffd@0 | 1 function state = sample_mdp(prior, trans, act) |
| wolffd@0 | 2 % SAMPLE_MDP Sample a sequence of states from a Markov Decision Process. |
| wolffd@0 | 3 % state = sample_mdp(prior, trans, act) |
| wolffd@0 | 4 % |
| wolffd@0 | 5 % Inputs: |
| wolffd@0 | 6 % prior(i) = Pr(Q(1)=i) |
| wolffd@0 | 7 % trans{a}(i,j) = Pr(Q(t)=j | Q(t-1)=i, A(t)=a) |
| wolffd@0 | 8 % act(a) = A(t), so act(1) is ignored |
| wolffd@0 | 9 % |
| wolffd@0 | 10 % Output: |
| wolffd@0 | 11 % state is a vector of length T=length(act) |
| wolffd@0 | 12 |
| wolffd@0 | 13 len = length(act); |
| wolffd@0 | 14 state = zeros(1,len); |
| wolffd@0 | 15 state(1) = sample_discrete(prior); |
| wolffd@0 | 16 for t=2:len |
| wolffd@0 | 17 state(t) = sample_discrete(trans{act(t)}(state(t-1),:)); |
| wolffd@0 | 18 end |