HMM - Compute probability that a given position in the sequence has an underlying state

[GoogleCodeExporter]

http://ai.stanford.edu/~serafim/CS262_2007/notes/lecture5.pdf
Page 12

Compute P(x1 ... xi, Statei = k)

Right now it seems that the API does not support this functionality.

Original issue reported on code.google.com by androidu...@gmail.com on 25 Aug 2014 at 2:44

[GoogleCodeExporter]

Thanks for submitting the issue! Indeed, it seems there is no current way to 
compute the probability of one state in the middle of a given sequence. This 
would be a nice functionality to add; and if someone would like to submit a 
patch for it, it would be welcome. If not, I will schedule that for some of the 
next releases. Thanks!

Original comment by cesarso...@gmail.com on 28 Aug 2014 at 2:41

• Changed state: Accepted

[GoogleCodeExporter]

New HiddenMarkovModel.Posterior() methods have been added to the 
HiddenMarkovModel and HiddenMarkovModel classes to compute posterior state 
probabilities as described in the aforementioned paper.

The issue has been tracked and closed on GitHub:
https://github.com/accord-net/framework/issues/31

Original comment by cesarso...@gmail.com on 9 Oct 2014 at 9:09

• Changed state: Started

[GoogleCodeExporter]

Available on Accord.NET 2.14

Original comment by cesarso...@gmail.com on 8 Dec 2014 at 9:14

• Changed state: Fixed