tensorflow / tensorflow/probability
Question about the generalization capability of a BNN
Nobody has claimed this yet.
- Dominant language
- Jupyter Notebook
- Stars
- 4.4k
- Forks
- 1.1k
- PR merge metrics
- No merged PRs in 30d
Description
I know that one of the main benefits of a BNN is the ability to gauge the confidence of a prediction. I understand why this can be important for various applications but i am more concerned about the advantages of BNN's ability to generalize over a traditional NN with the right selection of generalization techniques (batch norm, dropout, regularization, etc.).
Specifically i am wondering from a generalization perspective why BNN is better than just adding gaussian noise to our modeL?
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 issue names no files, tests, or entry points and asks a conceptual question about BNN generalization versus Gaussian noise. There is no implementation-defined completion condition; treat it as a request for an explanation rather than a first-contribution coding task.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- tensorflow
- Domain
- machine-learning
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 15/100