Compare the greenhouse population mix

A journey through what changed after the model was fitted? with Botie.

Choosing and Checking a Model · A journey with Botie

First visit? Learn the workshop controls

Botie's notebook

When Training and Use No Longer Share the Same Distribution

Part 1 of 2

A predictor's evaluation can stop describing its environment when the frequencies or relationships in incoming data change. Detecting that mismatch is a separate problem from fitting the original predictor.

Sources for this history

Lipton, Wang, and Smola: Detecting and Correcting for Label Shift (2018)

← Previous lesson: Does Eighty Percent Mean Eighty Percent?Read the full lesson and derivations