TITLE:

DATA ANALYSIS AND THE SINGULAR VALUE DECOMPOSITION (No. 145)

DATE:

Friday, March 16th, 2007

TIME:

3:30 PM

LOCATION:

GMCS 214

SPEAKER:

Gene Golub, Computer Science Deptartment, Stanford University

ABSTRACT:

In this talk, we describe a number of engineering and biological applications where the singular value decomposition (SVD) can be helpful in analyzing large data sets. The SVD was originally discovered in the 19th century but its use came into prominence with the advent of modern computers. Beginning in 1965, a series of algorithms were developed which made it practical to use the algorithm. With these algorithms, it is possible to determine low rank approximations to data sets, angles between subspaces and the pseudo-inverse and many other matrix functions. New methods continue to be developed for sparse data structures and for various computational environments. Our talk will present these ideas as well as some exciting new developments.

HOST:

Jose Castillo

DOWNLOAD: