tensorflow / tensorflow/probability
HiddenMarkovModel Binomial Extension
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
Hi,
I'm currently trying to implement a HMM model in which the observation distribution is binomial.
I see in the current tfp.Distributions.HMM the
observation distributionis A tfp.distributions.Distribution-like instance. The rightmost batch dimension indexes the distribution of each observation conditioned on the corresponding hidden state.
If the observation distribution is not only conditioned on the corresponding hidden state but also the current time-step, what would be the recommended way of implementing this under the tensorflow-probability framework?
Thanks for any suggestions.
BR,
Rlyu
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
The question concerns tfp.distributions.HMM and its observation distribution. Start by reviewing the HMM observation-distribution API and how it handles hidden-state and time-step conditioning. Done means documenting or establishing a supported implementation approach for binomial, time-dependent observations.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- tensorflow
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100