Riemann--Hilbert Problems, Computation and Universality
Tom Trogdon
Courant Institute of Mathematical Sciences, New York University
Date and time:
Friday, November 14, 2014 - 3:00pm
Location:
ECCR 245
Abstract:
Riemann--Hilbert (RH) problems provide a powerful and rigorous tool to study many problems in pure and applied mathematics. Problems in integrable systems and random matrix theory have been solved with the aid of RH problems --- including a proof of universality for unitary random matrix ensembles. RH problems can be approached numerically with applications to the numerical solution of PDEs and the Monte Carlo sampling of random matrix ensembles. The resulting methods are seen to have accuracy and complexity advantages over previously existing methods. In this talk, I will discuss this numerical methodology and the concept of universality, making connections to the statistical analysis of numerical algorithm runtimes and a model of neural computation.