Xiaodong Li

Xiaodong Li 

2015 - Now: Assistant professor in the statistics department of UC Davis

2013 - 2015: Postdoc in the statisitics department, Wharton school, University of Pennsylvania, working with Prof. Tony Cai.

2013: Ph.D of Mathematics, Stanford University, advised by Prof. Emmanuel Candes.

2008: Bachelor of Science with major in Mathematics, Peking University.

My CV and google citation page.

I am currently interested in methods and theory in both machine learning and statistics.

In machine learning, I am particularly interested in how to develop computationally-efficient and memory-efficient algorithms based on randomization and optimization, examples include memory-efficient kernel machines via random sampling, randomized seeding in clustering, etc. My other research interests include developing robust machine learning algorithms based on convex optimization, and the theoretical foundation of nonconvex optimization.

In statistics, I am now interested in statistical inferences in high-dimensional setup, particularly in low-assumption frameworks.

Theoretical analysis is an essential part in my research. Instead of establishing consistency or convergence rates under idealized models or strong assumptions, I am currently interested in exploring properties of learning/statistical methods in low-assumption or assumptionless frameworks to understand their power and limitation as much as possible.

I have been actively working in the following topics:

  • Unsupervised learning
  • Convex and nonconvex optimization
  • Network analysis
  • Imaging and signal processing



4109 Mathematical Scineces, Davis, CA, 95616

Phone: 650-223-1084

Email: xdgli AT ucdavis DOT edu