ADVERTISEMENT

Link to John Urschel's new math paper

Abstract cut-and-paste.

In this paper, we develop a cascadic multigrid algorithm for fast computation of the Fiedler vector
of a graph Laplacian, namely, the eigenvector corresponding to the second smallest eigenvalue. This
vector has been found to have applications in fields such as graph partitioning and graph drawing.
The algorithm is a purely algebraic approach based on a heavy edge coarsening scheme and pointwise
smoothing for refinement. To gain theoretical insight, we also consider the related cascadic multigrid
method in the geometric setting for elliptic eigenvalue problems and show its uniform convergence under
certain assumptions. Numerical tests are presented for computing the Fiedler vector of several practical
graphs, and numerical results show the efficiency and optimality of our proposed cascadic multigrid
algorithm.
 
Re: Urschel puts his affiliation down as PSU, psu email address. Cool. ***

I'm suspecting this was written while he was still a student at PSU. It usually takes a good deal of time before the paper is published.

This post was edited on 3/18 10:56 AM by uppermaclion
 
Submission date on arXiv is 1 December 2014. I wonder if he was working...

...on the final version of the manuscript while he was playing during the NFL season. That would be truly amazing, but the paper was submitted by the second author, so maybe the final edits were done by the co-authors and John as first author read and approved the final manuscript. Either way, he is a remarkable guy.
 
I found a couple math errors, other than that, pretty good job by John.....
3dgrin.r191677.gif
 
ADVERTISEMENT
ADVERTISEMENT