Precision

Precision answers the question: "Of all the items the model predicted as positive, how many actually were?" It is calculated as true positives / (true positives + false positives). High precision means the model rarely makes false positive errors — when it says something is positive, it's usually right. Precision is critical in applications where false positives are costly: spam filtering (falsely marking important emails as spam), content moderation (incorrectly removing benign content), and medical testing (unnecessary treatment). There is typically a trade-off between precision and recall — increasing one tends to decrease the other, governed by the classification threshold. The F1 score balances both metrics. In information retrieval, precision measures what fraction of returned results are relevant.