"Structured Matrix Computations: Algorithms and Applications"

Ming Gu, Professor, UC Berkeley, Department of Mathematics

January 14th (Friday), 3:00pm
Webb Hall 1100

Fast and reliable algorithms for strucutred matrix computations is a very active area of research in numerical linear algebra. In this talk, we briefly review the rich history in this area and discuss recent advances in structured matrix computations. In particular, we show how structured matrix techniques can be successfully exploited to develop superfast direct solvers and preconditioners for large classes of symmetric positive definite sparse matrices.

About Ming Gu, Professor, UC Berkeley:

Ming Gu received his Ph.D. in computer science from Yale University in 1993. He joined the Math Department at UCLA in 1996 and has been a Professor of Applied Math at UC Berkeley since year 2000. In the last several years, his work has been concentrated on structured matrix computation techniques.

In particular, he has been working on structured direct solvers and preconditioners for large sparse symmetric positive definite matrices, which arise in very diverse areas of science and engineering. Unlike their traditional counterparts, these structured methods only require nearly linear computational time and storage.

Hosted by: CCDC Seminar Series