Abstract

A sequence of recent works has led to the development of nearly-linear time solvers for linear equations in Directed Laplacians. Among other things, this gives the first nearly-linear time algorithm for computing the stable distribution of directed random walks. I'll talk about the overall approach to solving these equations, developed by Michael and his co-authors, and I'll outline how we combined it with approximate Gaussian elimination to get the first nearly-linear time solver for Directed Laplacians.

Video Recording