Conditional Wasserstein GAN-based Oversampling of Tabular Data for Imbalanced Learning

TL;DR

This paper introduces a conditional Wasserstein GAN-based oversampling method tailored for imbalanced tabular data with mixed variable types, improving minority class representation for better classification performance.

Contribution

It presents a novel GAN-based oversampling approach that effectively models complex tabular data with numerical and categorical features, incorporating an auxiliary classifier loss for improved downstream classification.

Findings

GAN-based oversampling outperforms traditional methods on multiple datasets.

The method effectively handles mixed numerical and categorical data.

Empirical results show improved classification metrics with GAN oversampling.

Abstract

Class imbalance is a common problem in supervised learning and impedes the predictive performance of classification models. Popular countermeasures include oversampling the minority class. Standard methods like SMOTE rely on finding nearest neighbours and linear interpolations which are problematic in case of high-dimensional, complex data distributions. Generative Adversarial Networks (GANs) have been proposed as an alternative method for generating artificial minority examples as they can model complex distributions. However, prior research on GAN-based oversampling does not incorporate recent advancements from the literature on generating realistic tabular data with GANs. Previous studies also focus on numerical variables whereas categorical features are common in many business applications of classification methods such as credit scoring. The paper propoes an oversampling method based on a conditional Wasserstein GAN that can effectively model tabular datasets with numerical and categorical variables and pays special attention to the down-stream classification task through an auxiliary classifier loss. We benchmark our method against standard oversampling methods and the imbalanced baseline on seven real-world datasets. Empirical results evidence the competitiveness of GAN-based oversampling.