Extrapolation

noun

say it: “ek-strap-uh-LAY-shun” /ɪkˌstræpəˈleɪʃən/

Estimating a value beyond the range of the known data, by assuming the pattern continues.

What is extrapolation?

Extrapolation is guessing for inputs outside the range the model learned from, such as a 110° day when training had none above 80°. A line keeps going as if nothing changes, and a tree repeats its last leaf, so neither guess can be trusted.

from sklearn.linear_model import LinearRegression
model = LinearRegression().fit([[60], [70], [80]], [20, 40, 60])   # days of 60° to 80°
print(model.predict([[150]]))   # [200.]: no one sells lemonade on a 150° day

In a sentence

  • “Asking about a 150-degree day is extrapolation: the model never saw anything above 80.”
  • “The tree's extrapolation was flat: every huge home got the same guess.”

Level: Journeyman · More AI Basics terms at this level

Updated