True Negative

noun

say it: “TROO NEG-uh-tiv” /ˈtru ˈnɛɡətɪv/

A case where a test or model says 'no' and the real answer is no.

What is a true negative?

A true negative is a no the model got right: it let through an example that really isn't the positive class, such as a normal payment. In scikit-learn's confusion matrix for two classes, it's the top-left count. When the positive class is rare, most guesses are true negatives.

actual = [1, 0, 0, 1, 1, 0]    # 1 = fraud
guessed = [1, 0, 1, 0, 1, 0]
print(sum(a == 0 and g == 0 for a, g in zip(actual, guessed)))   # 2: normal payments let through

In a sentence

  • “Because fraud is rare, nearly every payment was a true negative.”
  • “The coffee purchase the check let through was a true negative.”

Level: Apprentice · More AI Basics terms at this level

Updated