Chris Junchi Li, Ph.D.


Welcome! I am currently a researcher at Private Sector Company in Shanghai, China. Previously, I was a Visiting Scientist in the Department of Electrical Engineering and Computer Sciences, University of California, Berkeley.

My research interests encompass convex and nonconvex stochastic optimization, saddle-point (minimax) optimization, theoretical statistics, machine learning, stochastic processes, and the theory of reinforcement and deep learning. Recently, my work has focused on understanding stochastic approximation and optimization procedures in statistical machine learning for large-scale datasets. By leveraging both discrete-time and continuous-time perspectives, I have developed sharp convergence theories for new and existing algorithms in convex and nonconvex optimization. Success in this research would lead to significant advancements in the computational efficiency of current learning algorithms and the development of new algorithms with theoretical rates of convergence that match worst-case and instance-dependent lower bounds.

Please feel free to reach out if you are interested in any aspects of my research.

Services


Brief Vita


Contact Information


You are welcome to reach me at junchi.li@alumni.duke.edu or junchi.li.duke@gmail.com