— Final Report¶
Team:
1. Executive summary (150 words max)¶
2. Problem definition¶
- Question:
- Audience:
- Decision:
- Why it's interesting/hard: <1–2 sentences>
3. Data¶
- Source + license + retrieval method + date (table).
- Shape, columns, and any data-quality issues found.
- Cleaning summary: 3–5 key decisions with reasons (full log in the notebook).
4. Exploratory data analysis¶
- 3 key findings (claim → number → figure reference).
- One figure embedded with a caption (at minimum).
5. Method & modeling¶
- Framing: target, features, supervised/unsupervised, task type.
- Model(s) tried, split/CV setup, seeds.
- Why this model fits the question (1 paragraph).
6. Results & evaluation¶
- Metric(s) with exact numbers; baseline comparison.
- Table: model vs. baseline.
- Interpretation: what the numbers mean for the decision in §2.
- Limitations of the evaluation (small test set, imbalance, leakage risks).
7. Streamlit application¶
- What it lets a user do (widgets, panels).
- One screenshot or detailed panel description.
- Link to run instructions (README).
8. AI-assisted analysis (CLO-3)¶
- What AI assistance was used (PandasAI/Ollama/LLM prompt/agent), if any.
- One example: prompt → output → your verification → verdict.
- Why the verification step matters (2–3 sentences).
9. Ethics & reproducibility¶
- Data sheet: source, who collected it, what it doesn't contain (privacy), known limitations.
- Bias check: any group analyzed differently? Sample sizes?
- Reproducibility: environment pins, seeds, restart-safe notebooks, cache strategy.
- AI-use disclosure: tool + prompt + verification per
../assessment-plan.md.
10. Threats to validity (from the proposal, updated)¶
11. Conclusion & future work¶
- Restate the answer to the question (1–2 sentences).
- 2–3 concrete next steps if given another week.
12. Appendix¶
- Cleaning log (or link to notebook section).
- Viva preparation notes (optional).
- References/data sources.