`Two seminars by Mike Saunders:----------------------------------------------Joint CRM Applied Math and McGill CSE seminar---------------------------------------------Time: Friday, November 14, 2008, 2:30pmPlace: McGill, Burnside Hall 1205Title: 40 Years of Linear Algebra and Optimization at StanfordSpeaker: Michael Saunders         Systems Optimization Laboratory (SOL)         Stanford UniversityAbstract:I came to Stanford in 1967 as a very green graduate student(not in today's ecological sense).  Computer Science was a newdepartment, as was Operations Research.  The CS qualifyingexams allowed 3 out of 5 topics, including numerical analysis.Alan George and I obtained permission to take one of the ORexams.  Thus began a career of applying stable matrix methodsto numerical optimization (as pioneered by Gene Golub, PhilipGill, and Walter Murray).We trace the impact of Gene inviting numerous researchers toSerra House (including Chris Paige and Bruce Murtagh), as wellas George Dantzig's creation of the Systems Optimization Labin the OR Department, and Gene's founding of the SCCM Program.The talk includes some illustrations of the use of optimizationwithin the aerospace industry.Coffee and cakes are served after the seminar in the lounge roomon the 10th floor.---------------------------------------Michael will also give another talk:Time: Monday, Nov 17, 2008, 2:30pmPlace: McGill, McConnell Engineering Building, Room 103Title: Computing Approximate Pagerank Vectors by Basis Pursuit DenoisingAbstract:Basis Pursuit Denoising (BPDN) finds sparse solutions xto underdetermined systems Ax ~= b by balancing the1-norm of x against the 2-norm of the residual:   min_{x,r}  lambda||x||_1 + 1/2 r'r,   Ax + r = b.The PageRank eigenvector problem involves a square systemAx = b in which x is naturally nonnegative and somewhatsparse (depending on b).  We seek an approximate x that isnonnegative and extremely sparse.  We experiment with anactive-set optimization method designed for the dual ofthe BPDN problem, and find that it tends to extract theimportant elements of x in a greedy fashion._______________________________________________crmandcse mailing listcrmandcse@cs.mcgill.cahttp://mailman.cs.mcgill.ca/mailman/listinfo/crmandcse`