Generative Modelling for Structured Data

We study generative models for structured data such as tables and time series. Our work spans Generative Adversarial Networks (GANs), Large Language Models (LLMs) and Diffusion models, focusing on synthetic data quality, downstream utility, and privacy. We also explore federated and decentralized settings, which commonly arise in real-world collaborative scenarios.

Publications

1. WaveStitch: Flexible and Fast Conditional Time Series Generation with Diffusion Models
   • Conference: ACM SIGMOD 2026
   • Authors: Aditya Shankar, Lydia Y. Chen, Arie van Deursen and Rihan Hai

   • ��Paper

     ��Code

2. Federated Time Series Generation on Feature and Temporally MisalignedData
   • Conference: ECML PKDD 2025
   • Authors: Zhi Wen Soi, Chenrui Fan, Aditya Shankar, Abel Malan, Lydia Y. Chen

   • ��Paper

     ��Code

3. TabuLa: Harnessing Language Models for Tabular Data Synthesis
   • Conference: PAKDD 2025
   • Authors: Zilong Zhao, Robert Birke, and Lydia Y. Chen

   • ��Paper

     ��Code

4. GTV: Generating Tabular Data via Vertical Federated Learning
   • Conference: DSN 2023
   • Authors: Zilong Zhao , Han Wu, Aad van Moorsel, Lydia Y. Chen

   • ��Paper

     ��Code

5. SiloFuse: Cross-Silo Synthetic Data Generation with Latent Tabular Diffusion Models
   • Conference: ICDE 2024
   • Authors: Aditya Shankar, Hans Brouwer, Rihan Hai, Lydia Y. Chen
   • ��Paper
6. CTAB-GAN: Effective Table Data Synthesizing
   • Conference: ACML 2021
   • Authors: Zilong Zhao, Aditya Kunar, Robert Birke, Lydia Y. Chen

   • ��Paper

     ��Code