Scientific Computing and Matrix Computations Seminar

Organized by James Demmel and Ming Gu

Fall 2014

Wednesdays 12:10PM-1:00PM in 380 Soda Hall.


Past Seminars


Current Schedule








Date Speaker Affiliation Talk
Sept. 10, Jed Duersch
UC Berkeley
Variations on LOBPCG
Sept. 17
Andrew Gearhart
UC Berkeley
Bounds on the Energy Consumption of Computational Kernels
Sept. 24 Gregorio Malajovich
Universidade Federal do Rio de Janeiro
On Self-Convexity

Oct. 1 Lin Lin
UC Berkeley and LBNL
Finding the sparse representation of a dense matrix
Oct. 8
Laura Grigori
Inria
Enlarged Krylov subspace methods for reducing communication
Oct. 15
Grey Ballard
Sandia National Lab
How practical is fast matrix multiplication?
Oct. 22
Eugene Vecharynski
LBNL
A Projected Preconditioned Conjugate Gradient algorithm for computing a large invariant subspace of a Hermitian matrix
Oct. 29
Dave Anderson

UC Berkeley
An Efficient Algorithm for Spectral Graph Sparsification
Nov. 5
Chris Melgaard
UC Berkeley
Improved Column-Based Matrix Reconstruction
Nov. 12
Jiawang Nie
UC San Diego
Generating Polynomials and Symmetric Tensor Decompositions
Nov. 19
Lek-Heng Lim
U. of Chicago
Distances Between Subspaces of Different Dimensions
Nov. 26
Ming Gu
UC Berkeley
An efficient algorithm for computing a generalized Weaver partition
Dec. 3
Dan Spielman
Yale University

Dec. 10
Jianwei Xiao
UC Berkeley