What is distribution shift, and how do you detect it between train and production?
PICTURE THIS: DATA SPLIT
Fit on train, tune on val, report on test once.
Simple meaning
Distribution shift is a change in the input, label, or joint distribution after you ship.
WHY — EDA instead of guessing?
Why interviewers care about EDA:
who only read docs from people who shipped.
and tied to Data Science work.
Name the idea, why it exists, then one short example.
End with when you use it and one common pitfall.
STEPS — What happens step by step?
Before you speak the answer, walk the interviewer through these steps:
- 1Distribution shift is a
change in the input, label, or joint distribution after you ship.
- 2Compare covariate histograms, PSI
or KL-style scores, missingness rates, and outcome rates by time.
- 3When the label lags,
monitor input shift first so you are not surprised when the model decays.
- 4Give an example
One tiny concrete case you can say aloud.
- 5Common mistake
What juniors usually get wrong.
- 6Close
When you pick this over the alternative.
EXAMPLE — See it in action
Here's a short line you can speak, broken into clear beats:
Note: Adapt this scaffold to your own project — keep it under 60–90 seconds.
Key takeaway
Distribution shift is a change in the input, label, or joint distribution after you ship. Compare covariate histograms, PSI or KL-style scores, missingness rates, and outcome rates by time.