Metrics, Thresholds, and Error Analysis
Learners can choose metrics for a decision, compare thresholds, read confusion-matrix metrics, and write a model-quality note.
Executable courseRead the lesson pages, run the browser practice notebook, and keep the downloadable fallback pack if you want a local copy.
Module sequence
- Metrics follow the decisionUnit 1 · 20-35 minutes
- Regression metrics: MAE, RMSE, residual plots, and outlier sensitivityUnit 2 · 20-35 minutes
- Classification metrics: confusion matrix, accuracy, precision, recall, and F1Unit 3 · 20-35 minutes
- Class imbalance and why accuracy can misleadUnit 4 · 20-35 minutes
- ROC-AUC and PR-AUC as ranking summariesUnit 5 · 20-35 minutes
- Thresholds, costs, and choosing an operating pointUnit 6 · 20-35 minutes
- Calibration and uncertainty in simple languageUnit 7 · 20-35 minutes
- Error slices and subgroup checksUnit 8 · 20-35 minutes
- Project step: write a model-quality noteUnit 9 · 20-35 minutes
- Activity: Choose a Threshold and Explain ErrorsPractice activity · 45-75 minutes
- Module 06 Assessment: Metrics and Thresholds CheckModule check · 35-50 minutes