TAEGAN: Revisit GANs for Tabular Data Generation

Jiayu Li, Zilong Zhao, Kevin Yee, Uzair Javaid, Biplab Sikdar
Proceedings of the 17th Asian Conference on Machine Learning, PMLR 304:670-685, 2025.

Abstract

Synthetic tabular data generation has gained significant attention for its potential in data augmentation and privacy-preserving data sharing. While recent methods like diffusion and auto-regressive models (i.e., transformer) have advanced the field, generative adversarial networks (GANs) remain highly competitive due to their training efficiency and strong data generation capabilities. In this paper, we introduce Tabular Auto-Encoder Generative Adversarial Network (TAEGAN), a novel GAN-based framework that leverages a masked auto-encoder as the generator. TAEGAN is the first to incorporate self-supervised warmup training of generator into tabular GANs. It enhances GAN stability and exposes the generator to richer information beyond the discriminator’s feedback. Additionally, we propose a novel sampling method tailored for imbalanced or skewed data and an improved loss function to better capture data distribution and correlations. We evaluate TAEGAN against seven state-of-the-art synthetic tabular data generation algorithms. Results from eight datasets show that TAEGAN outperforms all baselines on five datasets, achieving a 27% overall utility boost over the best-performing baseline while maintaining a model size less than 5% of the best-performing baseline model. Code is available at: https://github.com/BetterdataLabs/taegan.

Cite this Paper

BibTeX
@InProceedings{pmlr-v304-li25c, title = {TAEGAN: Revisit GANs for Tabular Data Generation}, author = {Li, Jiayu and Zhao, Zilong and Yee, Kevin and Javaid, Uzair and Sikdar, Biplab}, booktitle = {Proceedings of the 17th Asian Conference on Machine Learning}, pages = {670--685}, year = {2025}, editor = {Lee, Hung-yi and Liu, Tongliang}, volume = {304}, series = {Proceedings of Machine Learning Research}, month = {09--12 Dec}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v304/main/assets/li25c/li25c.pdf}, url = {https://proceedings.mlr.press/v304/li25c.html}, abstract = {Synthetic tabular data generation has gained significant attention for its potential in data augmentation and privacy-preserving data sharing. While recent methods like diffusion and auto-regressive models (i.e., transformer) have advanced the field, generative adversarial networks (GANs) remain highly competitive due to their training efficiency and strong data generation capabilities. In this paper, we introduce Tabular Auto-Encoder Generative Adversarial Network (TAEGAN), a novel GAN-based framework that leverages a masked auto-encoder as the generator. TAEGAN is the first to incorporate self-supervised warmup training of generator into tabular GANs. It enhances GAN stability and exposes the generator to richer information beyond the discriminator’s feedback. Additionally, we propose a novel sampling method tailored for imbalanced or skewed data and an improved loss function to better capture data distribution and correlations. We evaluate TAEGAN against seven state-of-the-art synthetic tabular data generation algorithms. Results from eight datasets show that TAEGAN outperforms all baselines on five datasets, achieving a 27% overall utility boost over the best-performing baseline while maintaining a model size less than 5% of the best-performing baseline model. Code is available at: https://github.com/BetterdataLabs/taegan.} }
Endnote
%0 Conference Paper %T TAEGAN: Revisit GANs for Tabular Data Generation %A Jiayu Li %A Zilong Zhao %A Kevin Yee %A Uzair Javaid %A Biplab Sikdar %B Proceedings of the 17th Asian Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2025 %E Hung-yi Lee %E Tongliang Liu %F pmlr-v304-li25c %I PMLR %P 670--685 %U https://proceedings.mlr.press/v304/li25c.html %V 304 %X Synthetic tabular data generation has gained significant attention for its potential in data augmentation and privacy-preserving data sharing. While recent methods like diffusion and auto-regressive models (i.e., transformer) have advanced the field, generative adversarial networks (GANs) remain highly competitive due to their training efficiency and strong data generation capabilities. In this paper, we introduce Tabular Auto-Encoder Generative Adversarial Network (TAEGAN), a novel GAN-based framework that leverages a masked auto-encoder as the generator. TAEGAN is the first to incorporate self-supervised warmup training of generator into tabular GANs. It enhances GAN stability and exposes the generator to richer information beyond the discriminator’s feedback. Additionally, we propose a novel sampling method tailored for imbalanced or skewed data and an improved loss function to better capture data distribution and correlations. We evaluate TAEGAN against seven state-of-the-art synthetic tabular data generation algorithms. Results from eight datasets show that TAEGAN outperforms all baselines on five datasets, achieving a 27% overall utility boost over the best-performing baseline while maintaining a model size less than 5% of the best-performing baseline model. Code is available at: https://github.com/BetterdataLabs/taegan.
APA
Li, J., Zhao, Z., Yee, K., Javaid, U. & Sikdar, B.. (2025). TAEGAN: Revisit GANs for Tabular Data Generation. Proceedings of the 17th Asian Conference on Machine Learning, in Proceedings of Machine Learning Research 304:670-685 Available from https://proceedings.mlr.press/v304/li25c.html.

Related Material