tensorflow / tensorflow/probability
KL Divergence for Independent and Transformed distributions
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
It would be very helpful to have registered kl_divergence methods for independent and transformed distributions, as distribution b (ie allowing for the KL divergence between a Gaussian process and a Transformed MultivariateNormalLinearOperator). This would be especially helpful for transformed distributions with a mathematically trivial bijector, such as Reshape or Transpose. Is the list of registered kl_divergences going to be expanded in future?
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
No files, tests, or entry points are named. Start by reviewing the existing registered kl_divergence methods for Independent and Transformed distributions, then determine the supported distribution pairs and tests needed for the Gaussian-process and Transformed MultivariateNormalLinearOperator cases described here.
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