Bowen Li

Research

My research is centered on optimization theory and algorithms. I am particularly interested in accelerated and stochastic optimization, non-Euclidean optimization, and adaptive methods, with applications to scientific computing.

Stochastic Optimization

Nonlinear transformations of stochastic gradients, high-probability guarantees, and sharp complexity bounds under heavy-tailed noise.

Acceleration Techniques

Adaptive algorithms, splitting methods, convergence analysis, and continuous-time perspectives for first-order optimization.

Scientific Computing

Algorithms for structured optimization arising from scientific problems, with emphasis on problem-driven approaches and end-to-end complexity.