Stratified Sampling for Classification

Easy

machine-learning Microsoft Uber LinkedIn

Your dataset has 5% positive examples. After a random train/test split, the test set contains only 2% positive examples. How can you prevent this distributional shift?

Multiple Choice

Use a larger test set so the percentages balance out naturally Use stratified sampling to maintain the same class proportions in both splits Oversample the minority class in the test set only Shuffle the data multiple times before splitting

Correct!

Incorrect — try again next time!

Upgrade to Pro to unlock solutions and explanations

Stratified Sampling for Classification

Multiple Choice

Related problems

Deep-dive guides