Overview

SMOTE is a sophisticated over-sampling technique. Instead of simply duplicating existing minority instances, it creates new ones by selecting a minority instance and its nearest neighbors, then creating new points along the lines connecting them.

Benefits

  • Reduces the risk of overfitting compared to random over-sampling.
  • Provides more diverse data for the model to learn from.

Use Cases

  • Fraud detection.
  • Rare disease diagnosis.
  • Predicting equipment failure.

Related Terms