Andi Han

Andi Han is a Lecturer in Data Science at the School of Mathematics and Statistics, University of Sydney. He earned his PhD in Business Analytics from the University of Sydney Business School in 2023 and served as a postdoctoral researcher at RIKEN AIP’s Continuous Optimization Team until 2025.

• Research Interests:
  • Large generative models (diffusion models, large language models)
  • Optimization on manifolds
  • Efficiency of foundation models
  • Graph neural networks for biology and chemistry
• Awards:
  • DAAD AInet Fellowship (2025)
  • PhD Completion Award (USYD, 2023)
  • Best Paper Award (IEEE SCCI, 2022)
  • University Medal (USYD, 2019)
  • Business Analytics Prize (USYD, 2018)
• Teaching:
  • STAT5002: Introduction to Statistics (Unit Coordinator & Lecturer, S2 2025)
  • MATH1061: Mathematics 1A (Lecturer, S2 2025)

His recent publications focus on Riemannian optimization techniques, diffusion models, and graph neural networks (GNNs), with applications in protein sequence generation, transformer optimization, and AI for science. Collaborative work spans institutions like RIKEN, Zhejiang Lab, and A*STAR. He actively organizes workshops, including Deep Generative Model in Machine Learning: Theory, Principle and Efficacy at ICLR 2025.