mlsanityData to 7 Oct 2026

MargCTGAN: A "Marginally'' Better CTGAN for the Low Sample Regime

arXiv 2307.07997 ↗First posted 16 Jul 2023cs.LG cs.AIcited by 4

Not on the map. Why?

Abstract (arXiv)

The potential of realistic and useful synthetic data is significant. However, current evaluation methods for synthetic tabular data generation predominantly focus on downstream task usefulness, often neglecting the importance of statistical properties. This oversight becomes particularly prominent in low sample scenarios, accompanied by a swift deterioration of these statistical measures. In this paper, we address this issue by conducting an evaluation of three state-of-the-art synthetic tabular data generators based on their marginal distribution, column-pair correlation, joint distribution and downstream task utility performance across high to low sample regimes. The popular CTGAN model shows strong utility, but underperforms in low sample settings in terms of utility. To overcome this limitation, we propose MargCTGAN that adds feature matching of de-correlated marginals, which results in a consistent improvement in downstream utility as well as statistical properties of the synthetic data.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 18 of its 26 references that are in the map. First posted on arXiv 16 Jul 2023.

Builds on

18 of 26 references.

References · 18

  1. 2013Auto-Encoding Variational Bayescited by 3,980
  2. 2017Improved Training of Wasserstein GANscited by 2,223
  3. 2015Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networkscited by 2,048
  4. 2019Modeling Tabular data using Conditional GANcited by 695
  5. 2017Wasserstein GANcited by 635
  6. 2015A note on the evaluation of generative modelscited by 291
  7. 2017Generating Multi-label Discrete Patient Records using Generative Adversarial Networkscited by 197
  8. 2021How Faithful is your Synthetic Data? Sample-level Metrics for Evaluating and Auditing Generative Modelscited by 193
  9. 2018Data Synthesis based on Generative Adversarial Networkscited by 184
  10. 2021CTAB-GAN: Effective Table Data Synthesizingcited by 156
  11. 2020Synthetic Data -- Anonymisation Groundhog Daycited by 105
  12. 2017PacGAN: The power of two samples in generative adversarial networkscited by 104
6 more
  1. 2018Synthesizing Tabular Data using Generative Adversarial Networkscited by 92
  2. 2020Conditional Wasserstein GAN-based Oversampling of Tabular Data for Imbalanced Learningcited by 67
  3. 2020VAEM: a Deep Generative Model for Heterogeneous Mixed Type Datacited by 66
  4. 2022TabSynDex: A Universal Metric for Robust Evaluation of Synthetic Tabular Datacited by 27
  5. 2018Airline Passenger Name Record Generation using Generative Adversarial Networkscited by 19
  6. 2018RadialGAN: Leveraging multiple datasets to improve target-specific predictive models using Generative Adversarial Networkscited by 10

Built on by

Cited by year

Citing papers per year: 2024: 4.

4 papers.

  1. 2024Structured Evaluation of Synthetic Tabular Datacited by 7
  2. 2024DP-2Stage: Adapting Language Models as Differentially Private Tabular Data Generatorscited by 5
  3. 2024Towards Biologically Plausible and Private Gene Expression Data Generationcited by 2
  4. 2024High-Quality Tabular Data Generation using Post-Selected VAEcited by 0