Esc
↑↓ navigate   Enter open   Esc close

Data Science & ML

Data Leakage

/ˈdeɪtə ˈliːkɪdʒ/

Definition

Using information in model training that would not be available at prediction time, causing inflated performance metrics.

Example in context

"We included the total at checkout in training — it's derived from items bought, so it leaks the target label."

Related terms

  • Overfitting — When an ML model learns training data too precisely — including noise — and fails to generalise to new examples.

Practice this term

Master Data Leakage in context by working through exercises in the Data Science & ML module. You'll see the term used in real engineering scenarios with multiple-choice, fill-in-the-blank, and matching drills.

Practice in exercises → Browse all words

Frequently Asked Questions

What does "Data Leakage" mean?

Using information in model training that would not be available at prediction time, causing inflated performance metrics.

How do you pronounce "Data Leakage"?

"Data Leakage" is pronounced /ˈdeɪtə ˈliːkɪdʒ/.

Can you use "Data Leakage" in a sentence?

"We included the total at checkout in training — it's derived from items bought, so it leaks the target label."