Essential session storage is required to sign in. Optional analytics remain off unless you allow them.
Pick a real tabular dataset: build a leakage-safe sklearn Pipeline (split → preprocess → CV → held-out test), beat a stated baseline, track every run (MLflow or equivalent) with data/code versions, and write an evaluation report with intervals and a cost-aware metric choice.
No leakage (pipeline-fitted preprocessing, grouped/temporal splits where needed); test set touched once; intervals reported.
Dumb baseline stated and beaten; metric justified by error costs, not habit.
Runs tracked with params/data/code versions; a stranger can re-run the winner.
Defensible encodings; missingness handled deliberately; features exist at prediction time.
Use this README structure to make the work reviewable.
DevPath fetches a read-only, size-capped snapshot. Submitted code is never executed.
# End-to-end tabular ML pipeline ## Goal Pick a real tabular dataset: build a leakage-safe sklearn Pipeline (split → preprocess → CV → held-out test), beat a stated baseline, track every run (MLflow or equivalent) with data/code versions, and write an evaluation report with intervals and a cost-aware metric choice. ## Acceptance criteria - [ ] Evaluation rigor: No leakage (pipeline-fitted preprocessing, grouped/temporal splits where needed); test set touched once; intervals reported. - [ ] Baseline & metric: Dumb baseline stated and beaten; metric justified by error costs, not habit. - [ ] Reproducibility: Runs tracked with params/data/code versions; a stranger can re-run the winner. - [ ] Feature work: Defensible encodings; missingness handled deliberately; features exist at prediction time. ## 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.