Skip to content

Project Proposal

Team: <names> Date: <date> · Due W13 S26 · One page max

1. Question (one sentence)

We will find out: .

2. Audience & decision

  • Who cares:
  • The decision our answer would inform:

3. Dataset

Field Answer
Source
License
Retrieval
Shape (verified)
Verified loads? ✅ / ❌ (must be ✅ before submission)

4. Framing

  • Target variable:
  • Candidate features:
  • Type: supervised/unsupervised · regression/classification/clustering

5. Success metric (one checkable number)

Success =

6. Planned pipeline (3–4 bullets)

  1. Acquire + cache (data_loader.py).
  2. Clean + EDA (cleaning log, 3 figures).
  3. Model + evaluation (baseline comparison).
  4. Streamlit app (widgets: ).

7. Top risks & mitigations

Risk Mitigation

8. Threats to validity (4–6 sentences)

Under what conditions would your conclusion be wrong? Name at least one data problem (missingness, sampling bias) and one modeling problem (overfitting, leakage). This paragraph becomes part of the final report.

Approved: ______ (instructor)