Survey of resampling techniques for improving classification performance in unbalanced datasets

TL;DR

This survey reviews various resampling techniques designed to address class imbalance in datasets, analyzing their impact on classification performance to guide better model development.

Contribution

It provides a comprehensive overview of existing resampling methods and evaluates their effectiveness in improving classification outcomes on unbalanced datasets.

Findings

Resampling techniques can significantly improve minority class recall.

Some methods balance precision and recall effectively.

Performance varies depending on dataset and technique used.

Abstract

A number of classification problems need to deal with data imbalance between classes. Often it is desired to have a high recall on the minority class while maintaining a high precision on the majority class. In this paper, we review a number of resampling techniques proposed in literature to handle unbalanced datasets and study their effect on classification performance.