Try overweighting model errors when training
Open
Nobody has claimed this yet.
enhancement
- Dominant language
- Python
- Stars
- 570
- Forks
- 351
- Avg merge
- 2d 13h
- Merged PRs (30d)
- 65
Description
We currently consider all examples the same, without overweighting anything. We might want to try overweighting cases where the model made an error (e.g. identifying a bug as accessibility which was not actually accessibility).
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 training entry point. Start by locating the training pipeline and how model errors are identified, then clarify which errors should receive extra weight. Done should include an implemented weighting approach and evidence from training or evaluation that it improves classification.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- machine-learning, python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100