Practice · Foundational · Beginner

Precision and Recall

Precision is how many flagged items were correct; recall is how many of the real cases were caught.

What Precision and Recall is

The two trade off against each other. Loosening a threshold catches more true cases and more false alarms; tightening it does the reverse.

How it works

Precision is true positives divided by all positive predictions; recall is true positives divided by all actual positives. F1 combines them, and precision-recall curves show the whole trade-off across thresholds.

Why it matters

Choosing the operating point is a business decision, not a modelling one: a cancer screen and a marketing filter should sit in very different places on that curve.

Common uses

→Threshold selection
→Search relevance
→Content moderation
→Imbalanced classification

Strengths

✓Meaningful on imbalanced data where accuracy is useless

Watch for

✓Single F1 number hides which side you are failing on

Continue exploring

Confusion Matrix Classification