No observation at a step

[werner hoffmann]

Hello.

In my real world model it is possible, that I do not have a observation at a "time" step.

How would you model that?

I thought about:
-creating a new ObservationDiscrete WITHOUT
-hmm.setOpdf for this value to 0.00000000001

Only the transition matrix should be used for calculating the next step.

Is this correct?
Is there a more elegant way to do it?


Thanks for your help
Werner

[Jean-Marc François]

Hi,

There are quite a lot of academic papers dedicated to this subject, it's not an easy question.

If suits your problem, you could replace each sequence of your learning set containing a missing observation with N sequences where those missing observations have been replaced with a random observation.

Hth,
JM

--
You received this message because you are subscribed to the Google Groups "Jahmm HMM library" group.
To unsubscribe from this group and stop receiving emails from it, send an email to jahmm-discus...@googlegroups.com.
To post to this group, send email to jahmm-...@googlegroups.com.
Visit this group at http://groups.google.com/group/jahmm-discuss.
For more options, visit https://groups.google.com/groups/opt_out.