Essential session storage is required to sign in. Optional analytics remain off unless you allow them.
Pick a real tabular dataset (Kaggle/UCI) and build a supervised model end to end. Requirements: FRAME it as regression or classification with a clear target and success metric; do EDA and PREPROCESSING (handle missing values, scale numeric, encode categoricals) fit on the TRAINING split ONLY (no leakage); split train/test (or use cross-validation); train a baseline (linear/logistic regression) AND a second model (e.g. a random forest); EVALUATE on held-out data with metrics appropriate to the task (precision/recall/F1 for imbalanced classification, or MAE/RMSE for regression — not just accuracy); and interpret which features mattered. Explain one preprocessing or data-leakage decision you made.
Clear framing (target + metric); a train/test split or cross-validation; preprocessing fit on the training split only (no leakage).
A baseline plus a second model, evaluated on held-out data with metrics that fit the task (not just accuracy on imbalanced data).
Feature importance / interpretation of results, and a documented preprocessing-or-leakage decision.
Use this README structure to make the work reviewable.
DevPath fetches a read-only, size-capped snapshot. Submitted code is never executed.
# An end-to-end supervised ML project ## Goal Pick a real tabular dataset (Kaggle/UCI) and build a supervised model end to end. Requirements: FRAME it as regression or classification with a clear target and success metric; do EDA and PREPROCESSING (handle missing values, scale numeric, encode categoricals) fit on the TRAINING split ONLY (no leakage); split train/test (or use cross-validation); train a baseline (linear/logistic regression) AND a second model (e.g. a random forest); EVALUATE on held-out data with metrics appropriate to the task (precision/recall/F1 for imbalanced classification, or MAE/RMSE for regression — not just accuracy); and interpret which features mattered. Explain one preprocessing or data-leakage decision you made. ## Acceptance criteria - [ ] Correct ML setup: Clear framing (target + metric); a train/test split or cross-validation; preprocessing fit on the training split only (no leakage). - [ ] Modeling & evaluation: A baseline plus a second model, evaluated on held-out data with metrics that fit the task (not just accuracy on imbalanced data). - [ ] Insight: Feature importance / interpretation of results, and a documented preprocessing-or-leakage decision. ## 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.