Wei Liu

Research Assistant Professor
Department of Applied Mathematics
The Hong Kong Polytechnic University
Email: liuwei175 art lsec.cc.ac.cn
[Google scholar page], [Google homepage]
ORCID: 0000-0002-2376-8974

News

• 03.14: new paper accepted by Transactions on Machine Learning Research – "LoDAdaC: a unified local training‑based decentralized framework with Adam‑type updates and compressed communication".
• 01.26: new paper accepted by SIAM Journal on Optimization – "A SPIDER‑type stochastic subgradient method for expectation‑constrained nonconvex nonsmooth optimization".
• 01.06: new paper published by Mathematical Programming Computation – "Damped proximal augmented Lagrangian method for weakly‑convex problems with convex constraints".
• 07.09: new conference paper accepted by Transactions on Machine Learning Research – "Compressed decentralized momentum stochastic gradient methods for nonconvex optimization".

About me

I received my bachelor's degree in 2017 from the Department of Mathematics at Zhejiang University. I started my Master's study in September 2017 in the Academy of Mathematics and Systems Science of the Chinese Academy of Sciences under the supervision of Professor Xin Liu. I then became a PhD candidate there in August 2019. After that, I visited the Hong Kong Polytechnic University for two years, sponsored by Xiaojun Chen. I obtained the Ph.D. degree in June 2022. I was a Postdoctoral Research Associate at the Rensselaer Polytechnic Institute from 2022 to 2025, sponsored by Yangyang Xu. I now hold the position of Research Assistant Professor in the Department of Applied Mathematics at The Hong Kong Polytechnic University.

Selected publications

• Wei Liu, Yangyang Xu, A SPIDER‑type stochastic subgradient method for expectation‑constrained nonconvex nonsmooth optimization. SIAM JOURNAL on Optimization
• Wei Liu, Muhammad Khan, Gabriel Mancino‑Ball, Yangyang Xu, A stochastic smoothing framework for nonconvex‑nonconcave min‑sum‑max problems with applications to Wasserstein distributionally robust optimization.
• Hari Dahal, Wei Liu, Yangyang Xu, Damped proximal augmented Lagrangian method for weakly‑convex problems with convex constraints. Mathematical Programming Computation
• Wei Liu, Qihang Lin, Yangyang Xu, First‑order methods for affinely constrained composite non‑convex non‑smooth problems: Lower complexity bound and near‑optimal methods. Mathematics of Operations Research
• Wei Liu, Xin Liu, and Xiaojun Chen, An inexact augmented Lagrangian algorithm for training leaky ReLU neural network with group sparsity. Journal of Machine Learning Research
• Wei Liu, Xin Liu, and Xiaojun Chen, Linearly‑constrained nonsmooth optimization for training autoencoders. SIAM JOURNAL on Optimization

Research interests

• First‑order methods for large‑scale nonconvex (nonsmooth) constrained optimization.
• Stochastic (sub)gradient methods for statistical and machine learning.
• Complexity analysis (iteration, oracle and computational).
• Applications including fairness in AI, deep neural networks, distributed robust optimization, decentralized distributed learning, semi‑supervised learning, bilevel optimization, minimax problems and large language models.

Education

• Institute of Computational Mathematics and Scientific/Engineering Computing, Academy of Mathematics and Systems Science, China (2017 – 2022) — Ph.D. in Computational Mathematics (Advisor: Xin Liu).
• Zhejiang University, Zhejiang, China (2013 – 2017) — B.S. in Mathematics Pursuit Science Class, Chu Kochen Honors College (CKC College).

Experiences

• The Hong Kong Polytechnic University (Aug. 2019 – Aug. 2021) — Research Assistant, hosted by Professor Xiaojun Chen.
• Brown University (May 2023) — Workshop Visiting Scholar.
• Rensselaer Polytechnic Institute (Aug. 2022 – Aug. 2025) — Postdoctoral Research Associate, hosted by Professor Yangyang Xu.

Selected awards

• 2026 — Selected for the National High‑Level Young Talent Program.
• 2022 — CAS prize of president scholarship.
• 2017 — Outstanding Graduates in Zhejiang University (First Prize).
• 2014–2016 — Awarded three times on Basic Disciplines of Top‑notch Student Scholarship (First Prize).

Skills

• Programming languages: MATLAB, C, Python.
• Language: Chinese (Native), English (Professional working proficiency).
• Professional knowledge: Nonsmooth analysis, convex optimization, machine learning, first‑order methods in optimization.

Journal refereeing

1. Mathematics of Operations Research
2. Mathematical Programming Computation
3. Computational Optimization and Applications
4. Journal of Machine Learning Research
5. SCIENCE CHINA Mathematics
6. IEEE Transactions on Cybernetics
7. Journal of Computational Mathematics
8. Journal of Scientific Computing