Skip to content
BENIBIX Learn

Precision (Machine Learning)

AI📏 Evaluation & BenchmarksClassification Metrics

Also known as Precision, Positive Predictive Value

When a model says yes, how often it is right. Few false alarms means high precision.

Think of it like

It is like a fisher who keeps only real fish. Few old boots in the net means high precision.

Example

Emma's spam filter flags 200 emails. 190 really are spam and 10 are real mail, so its precision is 95%.

Why it matters

It tells you how far you can trust a yes. That matters when acting on a wrong yes hurts someone.

Definition

Precision is how often a model is right when it says yes. Out of all the cases it marked as yes, it is the share that really were yes. A spam filter with 95% precision is wrong only 5 times in 100 when it calls something spam. High precision means few false alarms. It matters most when a false alarm is costly, such as sending a real email to the spam folder.
Everything inside the circle was marked yes. Precision is the share inside that is real.

Common mistake

Pushing precision up and not watching recall. A model that says yes just once, and is right, has 100% precision but misses almost all.

Key takeaways

  1. 1
    Trust in a yes. Right when it says yes.
  2. 2
    Few false alarms. High precision.
  3. 3
    Pairs with recall. Watch both.

Compare