Synthesizing Tabular Data using Generative Adversarial Networks

By Lei Xu, Kalyan Veeramachaneni. 2018

323 citations recorded by the source corpus.

Generative adversarial networks (GANs) implicitly learn the probability distribution of a dataset and can draw samples from the distribution. This paper presents, Tabular GAN (TGAN), a generative adversarial network which can generate tabular data like medical or educational records. Using the power of deep neural networks, TGAN generates high-quality and fully synthetic tables while simultaneously generating discrete and continuous variables. When we evaluate our model on three datasets, we find that TGAN outperforms conventional statistical generative models in both capturing the correlation between columns and scaling up for large datasets.

Subfields: Computer Vision and Pattern Recognition, Statistical and Nonlinear Physics, Generative Models, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Stefan Lessmann
2. Ming Ding
3. Farhad Farokhi
4. L. Chen
5. Zilong Zhao
6. Vadim Borisov
7. Gjergji Kasneci
8. Bo Liu

In Statistical and Nonlinear Physics

1. Stefan Lessmann
2. Ming Ding
3. Justin Engelmann

In Generative Models

1. Stefan Lessmann
2. Ming Ding
3. L. Chen
4. Zilong Zhao
5. Vadim Borisov
6. Gjergji Kasneci
7. Robert Birke
8. Aditya Kunar

In Signal Processing

1. Stefan Lessmann
2. Ming Ding
3. Vadim Borisov
4. Gjergji Kasneci
5. Justin Engelmann
6. Tobias Leemann
7. Kathrin Sessler
8. Martin Pawelczyk

Citations in the corpus

• When Machine Learning Meets Privacy (2020, background): “[171] developed training deep neural networks for the generation of synthetic data that closely resemble the actual medical records of patients.”
• Deep Neural Networks and Tabular Data: A Survey (2021, background): “They often include missing values [34], extreme data (outliers) [67], and erroneous or inconsistent data [68] and have a small overall size relative to the high-dimensional feature vectors generated from the data [69].”
• CTAB-GAN: Effective Table Data Synthesizing (2021, methodology): “Beyond its success on generating images, GAN has recently been applied to generate tabular data [16, 19, 24, 25].”
• Conditional Wasserstein GAN-based Oversampling of Tabular Data for Imbalanced Learning (2020, methodology): “In the generator, crosslayers increase the variation of the noise, while in the discriminator and the AC they improve their discriminative power.”