Slides from some talks:
- Conference on Challenges in Geometry, Analysis, and Computation: High-Dimensional Synthesis.
Yale University, June 2012
Randomized Algorithms for Very Large-Scale Linear Algebra. - CU-Boulder "intro to research" seminar, Feb 2012
Fast methods for solving elliptic PDEs. - Applied mathematics seminar at Chalmers University, Jan 2012
Fast direct solvers for elliptic PDEs. - SIAM CSE conference in Reno, NV, Feb 2011
Fast direct solvers for elliptic PDEs. - Workshop on Modern Massive Data Sets, Palo Alto, June 2010
Randomized methods for computing the Singular Value Decomposition (SVD) of very large matrices. - CU-Denver Computational Mathematics Colloquium, Feb 2010
Fast numerical methods for solving elliptic PDEs. - Workshop on Random Matrix Methods, Banff, Jan 2010
Enabling very large-scale matrix computations via randomization. - Applied Mathematics seminar, Brown University, Jan 2010
Fast numerical methods for solving elliptic PDEs. - Workshop on Numerical Homogenization, Banff, Dec 2009
Fast and accurate techniques for computing Schur complements and performing numerical coarse graining. - NIPS Tutorial, Vancouver, Dec 2009
Randomization: Making Very Large Scale Linear Algebraic Computations Possible. - CU Boulder Applied Math Colloquium, Oct 2009
Fast matrix computations via randomized sampling. - Stanford applied math seminar, June 2009
Fast matrix computations via randomized sampling. - Lectures given in Sweden (Uppsala, KTH, Chalmers) in May 2009
Fast matrix computations via randomized sampling. - Georgiatech applied math seminar, April 2009
Fast numerical methods for solving linear PDEs. - AMS Annual Meeting, Jan 2009 (special session on the Mathematics of Knowledge and Information)
Fast matrix computations via randomized sampling. - GeorgiaTech, Dec 2008
Fast matrix computations via randomized sampling. - University of Toronto, April 2008
Fast matrix computations via randomized sampling. - University of Pennsylvania, Jan 2008
Fast numerical methods for solving linear PDEs. - University of Toronto, Jan 2008
Fast numerical methods for solving linear PDEs.
Randomized methods for the approximation of matrices (student talk). - IPAM, Lake Arrowhead workshop, Dec 2007
Randomized sampling of structured matrices. - Nov 2007, Duke applied math seminar.
Similar talks were given at Stanford, Berkeley, IPAM, and Caltech.
Fast numerical methods for solving linear PDEs. Berkeley version. IPAM version. - Los Angeles, Sep 2007, IPAM tutorial
Randomized methods for the approximation of matrices. - Zurich, July 2007, ICIAM
Two randomized methods for the approximation of matrices. - Zurich, July 2007, ICIAM
Fast direct solvers. - Zurich, July 2007, ICIAM
Rapid evaluation of electrostatic interactions in multi-phase media. - Austin, TX, March 2007, Numerical Analysis Seminar
Randomized algorithms and fast direct solvers. - Golden, CO, Dec. 2006. Math/CS Colloquium.
Methods for computing approximate solutions to linear boundary value problems. - Boulder, CO, Sep. 2006. Applied Math Colloquium.
An overview of numerical methods for linear PDEs, with exhortations to use methods based on integral equations. - Boulder, CO, Dec. 2005.
An informal talk describing my research to applied math grad students. - Boulder, CO, Oct. 2005.
The Fast Multipole Method. A talk in the computational math seminar. - Austin, TX, July 2005, USNCCM.
"Fast direct methods for boundary value problems". - Austin, TX, July 2005, USNCCM.
"Multiscale modelling of materials with periodic micro-structures". - Austin, TX, July 2005, USNCCM.
"Bandgap phenomena in materials with periodic skeletal micro-structures". - Stockholm, Uppsala, Goteborg, Kiel, Leipzig, and Lund, Feb. 2004.
"A fast direct solver for boundary integral equations in two dimensions".
- Karlstad, Feb 27th, 2004
"Boundary integral equation techniques for solving Partial Differential Equations".
- Austin, May 20th, 2004
"A fast direct solver for boundary integral equations in two dimensions".
- IABEM, May 24th, 2004
"A fast direct solver for boundary integral equations in two dimensions".