Module CA20
Data and model capstone and research communication
Demonstrate a reproducible model study and bounded operating plan with honest uncertainty.
Why this module
Demonstrate a reproducible model study and bounded operating plan with honest uncertainty.
What you produce
A defended model study and handover pack. Assess the output against the unit outcomes, explanation, safe handling, and independent transfer. Do not require public sharing or use of real sensitive records.
Chapters and units
CA20-C01: Problem and reproducibility
Define target, leakage risks, provenance, and permitted use; Record code, data, environment, and comparison conditions
- CA20-U01 Problem formulation and dataset acceptance Define target, leakage risks, provenance, and permitted use
- CA20-U02 Baselines and reproducible experiment Record code, data, environment, and comparison conditions
CA20-C02: Improvement and robustness
Test one justified change without contaminating final evaluation; Analyze slices and operational consequences
- CA20-U03 Model improvement and ablation Test one justified change without contaminating final evaluation
- CA20-U04 Robustness, fairness, and resource tradeoffs Analyze slices and operational consequences
CA20-C03: Operation and professional defense
Demonstrate a bounded scoring and response process; Explain limitations and reproduce the study on unfamiliar inputs
- CA20-U05 Serving or batch workflow and monitoring Demonstrate a bounded scoring and response process
- CA20-U06 Independent defense and handover Explain limitations and reproduce the study on unfamiliar inputs