What is class imbalance in classification problems?
Updated May 15, 2026
Short answer
Class imbalance occurs when some classes dominate the dataset.
Deep explanation
Imbalanced datasets bias models toward majority classes, reducing minority detection. Techniques like SMOTE, class weighting, and resampling help mitigate this issue.
Real-world example
Fraud detection where fraud cases are much rarer than normal transactions.
Common mistakes
- Using accuracy as the only metric in imbalanced datasets.
Follow-up questions
- What is SMOTE?
- Why is accuracy misleading here?