Bio

I am a fourth-year PhD student in Electrical and Computer Engineering (ECE) at the University of California, San Diego, advised by Prof. Yang Zheng. Previously, I received my bachelor degree in Bio-industrial Mechatronics Engineering, National Chung Hsing University, and master degree in ECE at the University of California, San Diego.

Research Interests

My research lies in continuous optimization, with a focus on developing principled and efficient algorithms for structured, nonsmooth, and constrained problems. I am broadly interested in understanding how problem geometry, regularity, and computational structure can be exploited to design adaptive optimization methods with strong theoretical guarantees. My work is motivated by applications in control, signal processing, and data science.

Updates

• [Sep. 2026] Check out our new work Revisiting Proximal Bundle Methods: Improved Rates under H ̈older Smoothness. Proximal bundle mehtods have long been considered difficult to analyze due to its elegant single-loop strucutre. In this paper, we revisit proximal bundle methods from a double-loop perspective, and show that their convergence rates automatically adapt to different levels of Hölder smoothness, without prior knowledge of the smoothness exponent or careful tuning of the step size. Our analysis is simple and modular. We hope that this perspective makes proximal bundle methods more accessible to a broader audience.

• [Sep. 2026] Check out our new work Weak Convexity and Proximal Bundle Methods for Nonsmooth Policy Optimization in Robust Control.

• [Dec. 2025] Check out our new work An accelerated proximal bundle method for convex optimization.

• [Sep. 2025] Check out our new work A Proximal Descent Method for Minimizing Weakly Convex Optimization.

• [Feb. 2025] Check out our new work A Bundle-based Augmented Lagrangian Framework: Algorithm, Convergence, and Primal-dual Principles.

• Our paper on Inexact Augmented Lagrangian Methods for Conic Programs: Quadratic Growth and Linear Convergence has been accepted to NeurIPS 2024. In this paper, we analyze the error bound condition for semidefinite programs (SDPs) from the perspective of exact penalty functions, and show linear convergence of the primal iterate of the Augmented Lagrangian method for SDPs.

• Presented a talk on Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods at 2024 ISMP conference.

• Our paper on Error bounds, PL condition, and quadratic growth for weakly convex functions, and linear convergences of proximal point methods has been accepted for presentation at the 2024 L4DC!

• Presented a talk on An Overview and Comparison of Spectral Bundle Methods for Primal and Dual Semidefinite Programs at 2024 IOS conference.

• Received the Informs 2023 poster presentation award - Operations Research and Optimization Methodologies Category Winner.

• Check out our new work An Overview and Comparison of Spectral Bundle Methods for Primal and Dual Semidefinite Programs.