massOfai

Precision

AI/ML Fundamentals

Of predicted positives, how many are actually positive

What is Precision?

Precision = True Positives / (True Positives + False Positives). Answers "When it predicts yes, how often is it right?"

Real-World Examples

  • Spam filter: of emails marked spam, 95% are actually spam
  • Medical test with low false positives

When to Use This

When false positives are costly (e.g., marking good emails as spam)