From mgu at math.berkeley.edu Sun Dec 4 23:41:07 2011 From: mgu at math.berkeley.edu (mgu@math.berkeley.edu) Date: Mon Dec 5 11:06:57 2011 Subject: [BANANA] Reminder: LAPACK seminar on Dec 07, Note change of time and venue In-Reply-To: <201112012014.pB1KEPGo022533@phoenix.math.berkeley.edu> References: <201112012014.pB1KEPGo022533@phoenix.math.berkeley.edu> Message-ID: <170bfb2c54ce79e27482a3b3b6b6a870.squirrel@calmail.berkeley.edu> Math 290, Section 25, CS 298, Section 6 Fall 2011 (Matrix Computations and Scientific Computing) Dec. 07 will be our last meeting for this semester. We will join hands with the Applied Math Seminar organized by Prof. Chorin to have a joint seminar at 4:10PM in 939 Evans Hall. For the schedule and other details about the seminar, please see math.berkeley.edu/~mgu/LAPACKSeminar.htm Date: Dec. 07, 2011 Speaker: Prof. James Bremer Title: Recent progress in integral equation methods Have a great winter break and see you back in January. From saunders at stanford.edu Mon Dec 5 10:36:25 2011 From: saunders at stanford.edu (Michael Saunders) Date: Mon Dec 5 11:07:37 2011 Subject: [BANANA] LA/Opt seminar Thursday Dec 8 (Chen Greif) Message-ID: Linear Algebra and Optimization Seminar (CME 510) http://icme.stanford.edu/seminars/seminars.php 4:15pm Thursday Dec 8, 2011 Y2E2 111 http://campus-map.stanford.edu/index.cfm?ID=04-070 Prof Chen Greif Dept of Computer Science The University of British Columbia, Vancouver, Canada greif@cs.ubc.ca http://www.cs.ubc.ca/~greif/ Multipreconditioned iterative methods for the solution of nonsymmetric linear systems of equations Standard Krylov subspace methods are widely used for the numerical solution of sparse and large linear systems, and their performance strongly depends on the availability of an effective preconditioner. Typically, these methods only allow the user to choose a single preconditioner, although in many situations there may be a number of possibilities. In this talk a new method is described, multi-preconditioned GMRES, which allows the use of more than one preconditioner. We give some theoretical results, propose a practical algorithm, and give some numerical results from problems in domain-decomposition and PDE-constrained optimization. From mgu at math.berkeley.edu Thu Dec 1 12:14:25 2011 From: mgu at math.berkeley.edu (Ming Gu) Date: Mon Dec 5 12:02:42 2011 Subject: [BANANA] LAPACK seminar on Dec 07, Note change of time and venue Message-ID: <201112012014.pB1KEPGo022533@phoenix.math.berkeley.edu> Math 290, Section 25, CS 298, Section 6 Fall 2011 (Matrix Computations and Scientific Computing) Dec. 07 will be our last meeting for this semester. We will join hands with the Applied Math Seminar organized by Prof. Chorin to have a joint seminar at 4:10PM in 939 Evans Hall. For the schedule and other details about the seminar, please see math.berkeley.edu/~mgu/LAPACKSeminar.htm Date: Dec. 07, 2011 Speaker: Prof. James Bremer Title: Recent progress in integral equation methods Abstract: Despite offering significant advantages over more direct approaches, integral equation methods are not widely used for the numerical solution of elliptic boundary value problems. This is primarily due to several unresolved issues which hamper the competitiveness of integral equation methods in certain situations. In this talk, I will describe several of these unresolved issues and discuss one in detail. In particular, I will focus on recent work regarding the numerical solution of integral equations given on domains with singularities. Solutions of integral equations arising from elliptic boundary value problems given on irregular domains often exhibit singularities which are more severe than those exhibited by the solution of the original equation. This is sometimes cited as a serious drawback of integral equation methods. In fact, as I will show, it turns out to be essentially harmless. If time permits, I will also discuss the efficient evaluation of a class of singular integrals which arises from the discretization of integral equations on surfaces. From mgu at math.berkeley.edu Thu Dec 1 12:15:36 2011 From: mgu at math.berkeley.edu (Ming Gu) Date: Mon Dec 5 12:02:42 2011 Subject: [BANANA] special seminar on numerical algebraic geometry (fwd) Message-ID: <201112012015.pB1KFaKj022543@phoenix.math.berkeley.edu> Speaker: Jonathan Hauenstein (Texas A&M University) Title: Numerical solving of polynomial equations: from 3264 & 1442 to 83200 & 38475 Time: 4-5pm Thursday December 8, 2011 Place: Evans 939 Abstract: Systems of polynomial equations arise in many areas of mathematics, science, economics, and engineering with their solutions, for example, describing equilibria of chemical reactions and economics models, and the design of specialized robots. Two classical problems, Steiner's problem of conics and Alt's problem of four-bar linkages, have significantly influenced the development of numerical methods used for solving polynomial systems, collectively called Numerical Algebraic Geometry. This talk will explore these classical problems using numerical algebraic geometric approaches. The talk will conclude with two problems related to the relationship of non-negatives polynomials and sums of squares where numerical algebraic geometry provides computational validation of results by Maulik and Pandharipande on Gromov-Witten and Noether-Lefschetz theory. From saunders at stanford.edu Mon Dec 5 23:28:29 2011 From: saunders at stanford.edu (Michael Saunders) Date: Tue Dec 6 06:23:29 2011 Subject: [BANANA] Monday: Dan Spielman gives Motwani Distinguished Lecture In-Reply-To: References: Message-ID: Skipped content of type multipart/alternative-------------- next part -------------- A non-text attachment was scrubbed... Name: Spielman.doc Type: application/msword Size: 40960 bytes Desc: not available Url : http://csmr.ca.sandia.gov/pipermail/banana/attachments/20111205/010d6789/Spielman-0001.doc From saunders at stanford.edu Thu Dec 8 09:40:37 2011 From: saunders at stanford.edu (Michael Saunders) Date: Thu Dec 8 09:41:26 2011 Subject: [BANANA] LA/Opt seminar TODAY (Chen Greif) Message-ID: Reminder: seminar this afternoon Linear Algebra and Optimization Seminar (CME 510) http://icme.stanford.edu/seminars/seminars.php 4:15pm Thursday Dec 8, 2011 Y2E2 111 http://campus-map.stanford.edu/index.cfm?ID=04-070 Prof Chen Greif Dept of Computer Science The University of British Columbia, Vancouver, Canada greif@cs.ubc.ca http://www.cs.ubc.ca/~greif/ Multipreconditioned iterative methods for the solution of nonsymmetric linear systems of equations Standard Krylov subspace methods are widely used for the numerical solution of sparse and large linear systems, and their performance strongly depends on the availability of an effective preconditioner. Typically, these methods only allow the user to choose a single preconditioner, although in many situations there may be a number of possibilities. In this talk a new method is described, multi-preconditioned GMRES, which allows the use of more than one preconditioner. We give some theoretical results, propose a practical algorithm, and give some numerical results from problems in domain-decomposition and PDE-constrained optimization. From saunders at stanford.edu Sun Dec 11 22:35:23 2011 From: saunders at stanford.edu (Michael Saunders) Date: Sun Dec 11 22:36:23 2011 Subject: [BANANA] Monday: Dan Spielman gives Motwani Distinguished Lecture In-Reply-To: References: Message-ID: Reminder to ICME LA/Opt community: Dan Spielman speaking Monday afternoon ---------- Forwarded message ---------- The first Rajeev Motwani Distinguished Lecture of the year will be given next Monday, December 12th, by Dan Spielman from Yale (title/abstract below). Dan is a legendary researcher and speaker, and you should definitely come. Motwani Distinguished Lectures are a roughly quarterly series of theory colloquia aimed at a relatively broad audience. The talk is at 11 AM on Monday December 12th in the CIS Auditorium ( http://campus-map.stanford.**edu/index.cfm?ID=04-050). There will be a reception with refreshments immediately after the talk. Hope to see you there! Tim --- Speaker: Dan Spielman (Yale) Title: Spectral Sparsification of Graphs and Approximations of Matrices Abstract: Expander graphs can be viewed as sparse approximations of complete graphs, with Ramanujan expanders providing the best possible approximations. We formalize this notion of approximation and ask how well a given graph can be approximated by a sparse graph. We prove that every graph can be approximated by a sparse graph almost as well as the complete graphs are approximated by the Ramanujan expanders: our approximations employ at most twice as many edges to achieve the same approximation factor. We also present an efficient randomized algorithm for constructing sparse approximations that only uses a logarithmic factor more edges than optimal. Our algorithms follow from the solution of a problem in linear algebra. Given any expression of a rank-n symmetric matrix A as a sum of rank-1 symmetric matrices, we show that A can be well approximated by a weighted sum of only O(n) of those rank-1 matrices. This is joint work with Joshua Batson, Nikhil Srivastava and Shang-Hua Teng. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://csmr.ca.sandia.gov/pipermail/banana/attachments/20111211/1376b467/attachment.html