Title of this talk:  The power of nonconvex optimization in solving random quadratic systems of equations

Abstract:  We consider the fundamental problem of solving random quadratic systems of equations in n variables, which spans many applications ranging from the century-old phase retrieval problem to various latent-variable models in machine learning. A growing body of recent work has demonstrated the effectiveness of convex relaxation --- in particular, semidefinite programming --- for solving problems of this kind. However, the
computational cost of such convex paradigms is often unsatisfactory, which limits applicability to large-dimensional data.

This talk follows another route: by formulating the problem into nonconvex programs, we attempt to optimize the nonconvex objectives directly. We demonstrate that for certain unstructured models of quadratic systems, nonconvex optimization algorithms return the correct solution in linear time, as soon as the ratio between the number of equations and unknowns exceeds a fixed numerical constant. We extend the theory to deal with noisy systems, and prove that our algorithms achieve a minimax optimal statistical accuracy. Numerical evidence suggests that the computational cost of our algorithm is about four times that of solving a least-squares problem of the same size.

This is joint work with Emmanuel Candes.

Bio: Yuxin Chen is currently an assistant professor in the Department of Electrical Engineering at Princeton University. Prior to joining Princeton, he was a postdoctoral scholar in the Department of Statistics at Stanford University, and he completed his Ph.D. in Electrical Engineering at Stanford University. His research interests include high-dimensional statistics, convex and nonconvex optimization, statistical learning, and
information theory. He received the 2019 AFOSR Young Investigator Award.