Lai TIAN

Lai Tian () Postdoctoral Scholar Daniel J. Epstein Department of Industrial & Systems Engineering University of Southern California Office: OHE 340 Email: tianlai.cs (at) gmail.com; laitian (at) usc.edu

About Me

I am a Postdoctoral Scholar at ISE of USC Viterbi working with Prof. Johannes O. Royset. Before relocating to Los Angeles, I spent four wonderful years in Hong Kong and received my Ph.D. from The Chinese University of Hong Kong (CUHK), where I was advised by Prof. Anthony Man-Cho So.

My research interests lie in the interface between mathematical optimization and computer science, with a particular emphasis on applications in modern machine learning. Recently, I have been especially focused on the computational and theoretical aspects of various approximation techniques in nonsmooth optimization and variational analysis.

Selected Recent Works [Google Scholar]

• Failure of uniform laws of large numbers for subdifferentials and beyond
  L. Tian, J. O. Royset
  Preprint, 2025. [arXiv]
  

• Approximating Rockafellians mitigate distributional perturbations: Discontinuous integrands and chance-constrained applications
  L. Tian, J. O. Royset
  Preprint, 2025. [arXiv]
  

• Testing approximate stationarity concepts for piecewise affine functions
  L. Tian, A. M.-C. So
  ACM-SIAM Symposium on Discrete Algorithms (SODA), 2025. [SIAM] [arXiv]
  Best Paper Prize for Young Researchers in Continuous Optimization, ICCOPT 2025

• No dimension-free deterministic algorithm computes approximate stationarities of Lipschitzians
  L. Tian, A. M.-C. So
  Mathematical Programming, Series A, 2024. [SpringerLink] [pdf]
  

• On the finite-time complexity and practical computation of approximate stationarity concepts of Lipschitz functions
  L. Tian, K. Zhou, A. M.-C. So
  International Conference on Machine Learning (ICML), 2022. [pdf]
  

Professional Service

• Review for Journals:

  • Applied Mathematics & Optimization

  • IEEE Signal Processing Magazine

  • IEEE Transactions on Image Processing

  • Journal of Global Optimization

  • Journal of Machine Learning Research

  • Mathematical Programming

  • Mathematics of Operations Research

  • SIAM Journal on Optimization

• Review for Conferences:

  • Neural Information Processing Systems (NeurIPS)

  • International Conference on Machine Learning (ICML)

  • International Conference on Learning Representations (ICLR)

  • International Conference on Artificial Intelligence and Statistics (AISTATS)

  • AAAI Conference on Artificial Intelligence (AAAI)

  • NeurIPS Workshop on Optimization for Machine Learning (OPT)

Misc.

• (Qiu-Qiu), (Wang-Zai), and (Ma-Tuan)