On Mar 8, 7:29 pm, Oscar Benjamin <oscar.j.benja...@gmail.com> wrote:
> On 8 March 2013 05:26,  <subhabangal...@gmail.com> wrote:
>
> > On Friday, March 8, 2013 2:18:06 AM UTC+5:30, subhaba...@gmail.com wrote:
>
> >> I was trying to learn Hidden Markov Model. In Python there are various 
> >> packages, but I was willing to do some basic calculation starting from the 
> >> scratch so that I can learn the model very aptly. Do you know of any thing 
> >> such?
>
> > My problem is over model validation on continuous time Markov system.
> > Generally, I understand the theory and can run the kits like HMM.py or 
> > Scikit-learn.
> > The problem is if I can not fit the data in run time I would be at the 
> > mercy of the kit.
> > So I wanted to know the coding of the computation.
>
> If those projects can do what you want then why not just look at their code?
>
> > I am specifically looking at the small python example of Forward, Backward 
> > and Viterbi
> > calculation. I tried to surf the web but did not help much.
>
> How about these two (both in 
> Python):http://en.wikipedia.org/wiki/Viterbi_algorithm#Examplehttp://en.wikipedia.org/wiki/Forward%E2%80%93backward_algorithm#Pytho...
>
> > I do not know much of Scientific forum.
>
> Try scipy-user as a start (someone there may know of a better mailing 
> list):http://mail.scipy.org/mailman/listinfo/scipy-user
>
> > I thought as HMM.py, NLTK, Scikit-learn are Python implementations so there 
> > would be lot
> > of people in the room who would know it.
> > And I got people like you, so I can not say I am wrong!
>
> Python as a programming language is used for many different things. I
> doubt that many people on this list will know what a Hidden Markov
> Model is. While I do know what they are, I have never done anything
> with them in Python or otherwise. However, if you ask in the right
> place, I'm sure that you will find people who have done similar things
> in Python.
>
> Oscar

Thank you Sir for your kind time. I would surely check them. Regards,
Subhabrata.
-- 
http://mail.python.org/mailman/listinfo/python-list

Reply via email to