Synthetic Data Generation

Fair Decision Making

Research on fair synthetic data generation and fair representation learning for bias-free decision making.

Time Series Generation

Advanced diffusion-based models for generating high-quality, long-sequence time series data.

Representative Synthetic Data for Fair Decision Making

TransFusion: Generating Long, High-Fidelity Time Series using Diffusion Models with Transformers

Fair4Free: Generating High-fidelity Fair Synthetic Samples using Data Free Distillation

Bt-GAN: Generating Fair Synthetic Healthdata via Bias-transforming Generative Adversarial Networks

Generative Models for Time-Series and Fair Data Generation

Generating Private and Fair Long-Sequenced Longitudinal Healthcare Records

Fair Latent Deep Generative Models (FLDGM) for Syntax-agnostic and Fair Synthetic Data Generation

Synthesizing Longer and Higher-Fidelity Time-Series Data Using Generative Adversarial Networks