📏  Precision (a.k.a. Positive Predictive Value, PPV)

Precision (a.k.a. Positive Predictive Value, PPV)#

The share of predicted positives that are truly positive.

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

What it is#

Precision (also positive predictive value, PPV) is the fraction of correct positive predictions among all positive predictions:

\[\text{Precision} = \frac{TP}{TP + FP}.\]

It answers: of everything the model flagged positive, how much really was?

The trade-off#

Precision counts false positives against you but ignores false negatives — so a model can reach high precision by only flagging the cases it is surest about. It is therefore read together with recall (\(TP/(TP+FN)\)), which counts the positives missed; the F1 score is their harmonic mean.

When it matters#

Precision is the priority when a false positive is costly — a spam filter deleting real mail, a system flagging innocent transactions as fraud — where you would rather miss some positives than raise false alarms.


Theme: Classification & Averaging Metrics  ·  All terminology



See also

Source article Adapted (context, re-expressed) in our own words from: Precision (a.k.a. Positive Predictive Value, PPV) (insightful-data-lab.com).

Tags: purpose: reference topic: terminology level: intermediate