stan-dev / stan-dev/projpred

Logit-normal or probit-normal as sampling distribution for `stat = "acc"`

Open
#503 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
R
Stars
114
Forks
31
PR merge metrics
No merged PRs in 30d

Description

For stat = "acc", it would probably make sense to assume a normal distribution as the sampling distribution for a transformed estimator (transformed from the interval $(0, 1)$ to $(-\infty, \infty)$, similarly to stat = "mse" and stat = "rmse" where the estimators are transformed from $(0, \infty)$ to $(-\infty, \infty)$ in #496), see the discussion at https://github.com/stan-dev/projpred/pull/496#discussion_r1739354823.

Note that stat = "R2" is not concerned by this as it may get negative, see https://github.com/stan-dev/projpred/pull/496#discussion_r1919204537.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the discussion linked from pull request #496 and review how the transformed sampling distributions for stat = "mse" and stat = "rmse" were handled. Decide whether acc should use a logit-normal or probit-normal distribution, then verify that R2 remains unaffected. Done means the chosen behavior is implemented and covered for stat = "acc".

Written by the indexing model from the issue text.

Assessment

Tech stack
r
Domain
machine-learning
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.