Knowledge Maxxing← Back to topic library
ML MaxxingBeginner60 sec
Precision
Precision is the proportion of a model's positive predictions that are actually correct, out of all the times it predicted positive.
Prerequisite: Accuracy
What matters
Precision matters most when a false positive is costly, since it directly measures how often the model's positive alarms turn out to be right.
Example
If a spam filter flags 100 emails as spam and 90 of them really are spam, its precision on that batch is 90%.
Common misconception
High precision does not mean the model catches most real positive cases; a model can have high precision while still missing many actual positives.
Recall question
What does high precision tell you about a model's positive predictions specifically?
Keywords and related concepts
precisionfalse positivesclassification metricrecall