Bi-Discriminator Class-Conditional Tabular GAN

TL;DR

This paper presents a bi-discriminator GAN tailored for generating realistic tabular data with mixed data types, improving distribution capture and data utility over existing methods.

Contribution

It introduces a novel bi-discriminator architecture with an adapted preprocessing scheme and conditional generator, enhancing synthesis quality for tabular datasets.

Findings

Outperforms existing models on benchmark datasets

Achieves higher likelihood fitness scores

Improves downstream machine learning performance

Abstract

This paper introduces a bi-discriminator GAN for synthesizing tabular datasets containing continuous, binary, and discrete columns. Our proposed approach employs an adapted preprocessing scheme and a novel conditional term for the generator network to more effectively capture the input sample distributions. Additionally, we implement straightforward yet effective architectures for discriminator networks aiming at providing more discriminative gradient information to the generator. Our experimental results on four benchmarking public datasets corroborates the superior performance of our GAN both in terms of likelihood fitness metric and machine learning efficacy.