Activity: Compare Model Families Honestly

Activity ID: ML-M08-A01 Estimated active time: 45-75 minutes

Goal

Produce the main Module 08 evidence: Model-comparison report with baseline, candidates, metric table, selected model, and rejected alternatives.

Inputs

• data/module_08_model_comparison_v1.csv
• data/data-dictionary.md
• notebooks/module-08-worked.ipynb
• notebooks/module-08-practice.ipynb

Instructions

1. Fit at least three model families on identical splits.
2. Record accuracy, training time, and one interpretability note for each.
3. Include both baselines from Module 3 in the same table.
4. Identify which model you would deploy and state the trade-off you accepted.
5. Change the random seed and refit. Record how much the ranking moves.
6. Write two limitations, including what the seed sensitivity implies.

Quality checklist

• [ ] All models share the same split.
• [ ] The table includes cost and interpretability, not accuracy alone.
• [ ] Baselines appear in the comparison.
• [ ] The deployment choice names its trade-off.
• [ ] Seed sensitivity was measured, not assumed.

Submission format

Submit the notebook output or worksheet table plus a 4-6 sentence explanation.

Safety note

Use only supplied fictional or synthetic data. Do not add or upload real learner data.