Tuesday July 26/10:30/Grande Ballroom

Invited Presentation 4

Chair: James McKenna, Bellcore

Computation with Markov Chains

The long term behavior of a Markov chain is studied by determining the steady state vector. The computation of this vector can be cast as a matrix eigenvector or nullspace problem. The speaker will survey the state-of-the-art, highlighting two new methods: a block form of the GTH algorithm that is highly effective for dense matrix models, and an iterative method that gives good results on sparse parametric models. She will conclude with a discussion of application areas and research opportunities.

Dianne P. O'Leary, Computer Science Department, University of Maryland, College Park

Dianne O'Leary is a professor of Computer Science at the University of Maryland, College Park. She received her Ph.D. from Stanford University in 1976. She is currently a member of the University of Maryland Institute for Advanced Computer Studies and a consultant at the National Institute of Standards and Technology. Her research interests are in the areas of numerical linear algebra and optimization, ill-posed problems, regression, Markov chains, and signal processing. She has served on the SIAM Council and on the editorial boards of several SIAM journals, and is active in the Association for Women in Mathematics. She was recently named a Distinguished Alumna by the School of Science of Purdue University.