scikit-learn / scikit-learn/scikit-learn
IterativeImputer shouldn't just use l2 loss by default
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 67.3k
- Forks
- 27.4k
- Avg merge
- 1d 15h
- Merged PRs (30d)
- 58
Description
@GaelVaroquaux points out that iterative imputation with a regularised least-squares model is more-or-less the same as using NMF for imputation. We should instead use RandomForestRegressor as the default regressor in IterativeImputer, at least if sample_posterior=False (or we can implement predict(return_std=True) on RandomForestRegressor!).
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
Start by reading the IterativeImputer entry point and the behavior around sample_posterior=False. Compare the current regularised least-squares default with RandomForestRegressor, including whether posterior sampling requires predict(return_std=True). Done means the default choice and its posterior-sampling behavior are defined and covered by tests.
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
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100