1. Modeling — Three Required Approaches ⚠️

7/7

✓	M1	🤖 Naive baseline (majority class / mean predictor)
✓	M2	🤖 Classical (non-DL) ML model
✓	M3	🤖 Deep learning model
✓	M4	👤 All three documented in README with file locations
✓	M5	👤 Rationale for each model written up in report
✓	M6	👤 Selected final model clearly identified (and justified)
✓	M7	🤖 Trained model artifacts present or reproducible

2. Required Experimentation ⚠️

6/6

✓	EX1	👤 — Experiment is well-motivated: poses a specific question about your system
✓	EX2	👤 — Experimental plan documented (hypothesis, method, metrics)
✓	EX3	👤 — Results reported with numbers/visualizations
✓	EX4	👤 — Interpretation written: what do the results tell you?
✓	EX5	👤 — Experiment directly informs a modeling or system design decision
✓	EX6	👤 — Actionable recommendations drawn from experiment

3. Interactive Application ⚠️

5/5

✓	APP1	🤖 — App exists and runs inference only (no training in app code)
✓	APP2	👤 — Good UX — polished interface, not a bare Streamlit demo
✓	APP3	👤 — Publicly accessible via internet (deployed URL)
✓	APP4	👤 — Live for at least one week after submission date (due 2026-04-14)
✓	APP5	🤖 — Deployment URL recorded in README.md

4. Written Report ⚠️

17/17

✓	R01	👤 — Problem Statement
✓	R02	👤 — Data Sources (with provenance and access method)
✓	R03	👤 — Related Work (review of prior literature)
✓	R04	👤 — Evaluation Strategy & Metrics (with justification — *"this is critical"*)
✓	R05	👤 — Modeling Approach → Data Processing Pipeline (with rationale per step)
✓	R06	👤 — Hyperparameter Tuning Strategy
✓	R07	👤 — Models Evaluated: Naive baseline, Classical ML, Deep learning (with rationale)
✓	R08	👤 — Results: quantitative comparison across all models and metrics
✓	R09	👤 — Results: visualizations and confusion matrices
✓	R10	👤 — Error Analysis: 5 specific mispredictions identified
✓	R11	👤 — Error Analysis: root cause explained for each
✓	R12	👤 — Error Analysis: concrete, specific mitigation strategies per case
✓	R13	👤 — Experiment Write-Up (plan → results → interpretation → recommendations)
✓	R14	👤 — Conclusions
✓	R15	👤 — Future Work ("what would you do with another semester?")
✓	R16	👤 — Commercial Viability Statement
✓	R17	👤 — Ethics Statement

5. In-Class Pitch (5 min hard stop) ⚠️

5/5

✓	P1	👤 — Problem & Motivation slide(s) *(Mike — slide deck)*
✓	P2	👤 — Approach Overview slide(s) *(Mike — slide deck)*
✓	P3	👤 — Live Demo prepared and rehearsed *(Mike — slide deck)*
✓	P4	👤 — Results, Insights, or Key Findings slide(s) *(Mike — slide deck)*
✓	P5	👤 — Presentation stays within 5 minutes *(Mike — slide deck)*

6. Code & Repository

26/27

✓		🤖 README.md — project description, setup instructions
✓		🤖 requirements.txt — all dependencies pinned
✓		👤 setup.py — data acquisition / project setup pipeline
✓		👤 main.py — entry point / app launcher
✓		🤖 scripts/make_dataset.py
✓		🤖 scripts/build_features.py
✓		🤖 scripts/model.py
✓		🤖 models/ — directory exists (artifacts generated after training)
✓		🤖 data/raw/, data/processed/, data/outputs/
✓		🤖 notebooks/ — directory exists
✓		🤖 .gitignore
✓		👤 REPO1 — At least one exploration notebook in notebooks/
✓	CQ1	🤖 — All code modularized into classes/functions (no loose executable code)
✓	CQ2	🤖 — No executable code outside if __name__ == "__main__" guards
✓	CQ3	👤 — Descriptive variable names throughout
✓	CQ4	👤 — Docstrings on all public functions
✓	CQ5	🤖 — AI usage attributed at top of each file that used AI assistance (link to source required)
✓	CQ6	👤 — External code/libraries attributed at top of relevant files
✓	GIT1	🤖 — Feature branches in use
✓	GIT2	🤖 — PR template in .github/PULL_REQUEST_TEMPLATE/
✓	GIT3	👤 — All code merged via PRs (no direct commits to main) — enforced by branch protection
✓	GIT4	🤖 — Every PR has a meaningful Summary — enforced by PR Summary CI check
○	GIT5	👤 — Every PR reviewed with substantive comments — CODEOWNER approval required
✓	GIT6	🤖 — .env is never committed (check .gitignore) — enforced by Secret Scan CI check
✓	GIT7	🤖 — Large data files / model binaries are never committed (50 MB limit) — enforced by Large File Scan CI check
✓	NOV1	👤 — Working on a dataset/problem with no prior modeling approaches, OR
✓	NOV2	👤 — Clearly explains what is new/novel vs. prior approaches (with citations), achieves near-SOTA or better explainability

7. Pre-Submission Checklist

7/9

✓		👤 FS1: python setup.py runs end-to-end without errors
✓		👤 FS2: python scripts/build_features.py produces processed data
✓		👤 FS3: python scripts/model.py trains all three models and writes artifacts to models/
✓		👤 FS4: App launches and is reachable in browser
✓		👤 FS5: Live deployment URL is working and accessible without login
✓		👤 FS6: README deployment link updated to actual URL
○		👤 FS7: Written report submitted in the required format
○		👤 FS8: Pitch deck/slides prepared
✓		👤 FS9: Repo is public (or access granted to grader)