Essential session storage is required to sign in. Optional analytics remain off unless you allow them.
Pick a public dataset and write a clean, reproducible notebook: explore the data (distributions, missingness, leakage risks), state a dumb baseline, train one simple model with a proper train/test split, and honestly report whether you beat the baseline and why. No tuning — the point is rigor, not score.
Split before any fitting; test set touched once; obvious leakage risks named and avoided.
A trivial baseline stated and compared against; result reported even if the model loses.
Distributions and missingness examined; findings drive the modeling choices.
Seeded; runs top-to-bottom on a clean kernel; dependencies pinned.
Use this README structure to make the work reviewable.
DevPath fetches a read-only, size-capped snapshot. Submitted code is never executed.
# EDA and a baseline model notebook ## Goal Pick a public dataset and write a clean, reproducible notebook: explore the data (distributions, missingness, leakage risks), state a dumb baseline, train one simple model with a proper train/test split, and honestly report whether you beat the baseline and why. No tuning — the point is rigor, not score. ## Acceptance criteria - [ ] Leakage-safe evaluation: Split before any fitting; test set touched once; obvious leakage risks named and avoided. - [ ] Honest baseline: A trivial baseline stated and compared against; result reported even if the model loses. - [ ] EDA quality: Distributions and missingness examined; findings drive the modeling choices. - [ ] Reproducibility: Seeded; runs top-to-bottom on a clean kernel; dependencies pinned. ## Verification - [ ] Document installation and run commands. - [ ] Record automated checks and their exact commands. - [ ] Add representative output, screenshots, or a short demo where useful. - [ ] Confirm failure paths and known constraints. ## Decision log ### Decision title - Context: - Choice: - Alternatives considered: - Tradeoffs: ## Evidence - Link each rubric criterion to the file, test, or artifact that demonstrates it. ## Known limitations - List what is intentionally out of scope and what should be improved next.