- #1
Srecko
- 11
- 0
Hi all,
Not sure if this would be the right place for this question, but I know it bothers me for some time already and would really appreciate any kind of help. I am trying to fit an HMM, but here for every observation in the sequence I have feature vector - probability distribution that given observation appears in different groups.
I have seen this approach http://www.cs.jhu.edu/~mpaul/files/emnlp2012-m4.pdf that basically uses log-linear function to translate a vector of continuous values into a discrete value. However, I'm not sure whether I understood it correctly nor how to apply it. My idea was to apply this (or similar) algorithm on the data I have and then use any existing HMM implementation (probably in R) to fit a model.
Thanks for any ideas, suggestions...
Not sure if this would be the right place for this question, but I know it bothers me for some time already and would really appreciate any kind of help. I am trying to fit an HMM, but here for every observation in the sequence I have feature vector - probability distribution that given observation appears in different groups.
I have seen this approach http://www.cs.jhu.edu/~mpaul/files/emnlp2012-m4.pdf that basically uses log-linear function to translate a vector of continuous values into a discrete value. However, I'm not sure whether I understood it correctly nor how to apply it. My idea was to apply this (or similar) algorithm on the data I have and then use any existing HMM implementation (probably in R) to fit a model.
Thanks for any ideas, suggestions...