Overview

Precision answers the question: 'Of all the instances the model predicted as positive, how many were actually positive?' It is also known as 'Positive Predictive Value.'

Formula

Precision = True Positives (TP) / (True Positives (TP) + False Positives (FP))

Use Case

Precision is critical when the cost of a 'False Positive' is high, such as in spam detection (you don't want to mark an important email as spam) or legal sentencing.

Related Terms