0

I have trained HMM for each class separately, and selected a highest likelihood model given the observation sequence.

However, my requirement is to build single HMM for all classes. Each class can have multiple hidden states from S1,....Sn. For example, if have three classes, then the total number of hidden states are 3xn. How do i train this kind of classifier?.

user1023514
  • 101
  • 1
  • 1
  • 4
  • What about training a mixture model for each class, assigning each mixture model to a state (or splitting the mixture over couple of states) and use these values as the initial values to train the HMM? Just an idea... I can elaborate if you feel the need. – Eskapp Feb 21 '17 at 20:56
  • yes, if you elaborate more, that would be great help. My data is a continuous stream. Input sequences for each class can be a variable length. If mixture model helps in deriving the inter class/sequence relations, its good to use mixture models as you suggested. – user1023514 Feb 22 '17 at 02:13

0 Answers0