What is data leakage in a train and test setup?
PICTURE THIS: DATA SPLIT
Fit on train, tune on val, report on test once.
Simple meaning
Leakage is any path that lets information from the evaluation set, or from the future, into training.
WHY — Train/Val/Test instead of guessing?
Why interviewers care about Train/Val/Test:
who only read docs from people who shipped.
and tied to AI / ML 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:
- 1Leakage is any path
that lets information from the evaluation set, or from the future, into training.
- 2Examples include scaling on
all rows, using the target to build a feature, and random splits of duplicated customers.
- 3Leaked models look brilliant
in notebooks and fail in production.
- 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
Leakage is any path that lets information from the evaluation set, or from the future, into training. Examples include scaling on all rows, using the target to build a feature, and random splits of duplicated customers.