Yuchen Lou

Ph.D candidate 2021–2026 (expected) Department of Industrial Engineering and Management Sciences Northwestern University

Contact

Email: yuchenlou2026[at]u.northwestern.edu

Google Scholar    LinkedIn    GitHub

About Me

Hi! I am Yuchen Lou (楼昱辰), a fourth-year Ph.D. candidate at Northwestern IEMS. I am fortunate to be advised by Prof. Jorge Nocedal and Prof. Andreas Waechter.

Before coming to Northwestern, I completed my Bachelor's degree in Mathematics from The University of Hong Kong (HKU) in 2021 with First Class Honors. I also spent a rewarding year working in the Department of Mathematics at UCLA.

My research centers on the theoretical and algorithmic foundations of Computational Optimization, particularly for problems arising in deep learning, scientific computing, and operations research. The long-term goal of my research is to bridge the gap between theoretical advances, practical solver implementations, and the complex requirements of real-world problems. A guiding principle in my work is captured well by Fletcher:

“The subject of optimization is a fascinating blend of heuristics and rigour, of theory and experiment.”
— R. Fletcher

Motivated by this philosophy, my research aims to develop optimization methods that are robust, structure-exploiting, and scalable, while grounded in illuminating and explanatory theory. I believe the best methodology emerges from deeply understanding the underlying structures and limitations of a problem, guided by optimization fundamentals and supported by systematic empirical validation.

Here are some topics I am exploring at the current stage:

• Structure-aware neural network training methods such as Shampoo

• Memory-efficient optimization for LLM

• Nonconvex bilevel, two-stage, and min-max optimization with constraints

• Noise-robust and derivative-free methods in deep learning and scientific computing

Research

Papers

• A Decomposition Framework for Nonlinear Nonconvex Two-Stage Optimization
  Yuchen Lou, Xinyi Luo, Andreas Waechter, Ermin Wei.
  Preprint submitted.

• Design Guidelines for Noise-Tolerant Optimization with Applications in Robust Design
  Yuchen Lou, Shigeng Sun, Jorge Nocedal.
  SIAM Journal on Scientific Computing, 2025.

• Globally Solving Concave Quadratic Program via Doubly Nonnegative Relaxation
  Zheng Qu, Tianyou Zeng, Yuchen Lou.
  Mathematical Programming Computation, 2025.

• A Zeroth-Order Block Coordinate Descent Algorithm for Huge-Scale Black-Box Optimization
  HanQin Cai, Yuchen Lou, Daniel McKenzie, Wotao Yin.
  ICML, 2021.

Talks

1. 2025 International Conference on Continuous Optimization, Jul. 2025, Los Angeles CA

2. 25th International Symposium on Mathematical Programming, Jul. 2024, Montreal Quebec

3. 2024 INFORMS Optimization Society Conference, Mar. 2024, Houston TX

4. 2023 INFORMS Annual Meeting, Oct. 2023, Phoenix AZ

5. 2021 INFORMS Annual Meeting, Oct. 2021, Anaheim CA (Virtual)

6. ICML 2021 Virtual Talk, Jun. 2021, Virtual

7. Research Colloquium of HKU Science Undergraduate Research, Nov. 2020, Hong Kong SAR

8. Zeroth Order Online Meeting, Jul. 2020, UCLA & Alibaba DAMO Academy (Virtual)

Miscellaneous

During free time I play (both mobile and arcade) rhythmic games, which (to some degree) explains my favor in EDM, HDM and jazz. My favorite band is Tokyo Incidents.