Confusion Matrix and Basic Metrics
How to extract and interpret TP, TN, FP, FN from a 2×2 table; why accuracy alone misleads on imbalanced data
Precision and Recall
Two metrics that expose what accuracy hides—false alarms vs missed cases, and the tradeoff between them
F-Scores and Balanced Metrics
How to balance precision and recall into a single number; when to use F1, F-beta, or alternatives like MCC
ROC, AUC, and Threshold Tuning
Visualising and comparing classifiers via ROC curves and AUC; how to pick a threshold when the default 0.5 isn't right
Multiclass and Choosing Metrics
Extending binary metrics to >2 classes; decision guide for which metric to use