Precision, Recall, and F1: Metrics Built for Imbalance

Reading a confusion matrix's four counts, the precision/recall/F1 formulas and what each one answers, why precision and recall trade against each other by threshold, and a real precision-recall curve on a severe 5 percent positive-class task.

~5 min total·4 quadrants of structured content

By the end of this topic, you will

Define true positive, false positive, true negative and false negative, and compute precision, recall and F1 from them
Explain why lowering a decision threshold raises recall while lowering precision, and why no threshold maximises both
Read a precision-recall curve and explain why it is more honest than ROC on severely imbalanced data

Quadrant 1 · e-Tutorial

Video lectures and walkthroughs

Video content coming soon.

Quadrant 2 · e-Content

Articles and case studies

Precision, Recall, and F1: Metrics Built for Imbalance

Precision asks: of everything flagged malicious, how much actually was? Recall asks: of everything actually malicious, how much did the model catch? A real Random Forest's precision-recall curve on this unit's own 5 percent task scored an average precision of 0.54, against a random classifier's 0.05.