Year: 2012
Type: dissertation
Abstract: Multiple orthogonal polynomial (MOP) ensembles form a generalization of the orthogonalpolynomial (OP) ensembles appearing in random matrix theory. Of maininterest is the asymptotic behavior of MOP ensembles, i.e., the behavior as the numberof particles tends to infinity.We consider the subclass of Angelesco ensembles. In these ensembles the particlesare separated into different groups, with both internal interaction within the groups,and mutual interaction between the groups. By tuning the parameters in the modelwe see a phase transition appear. This phase transition gives rise to critical behavior:under appropriate scaling the Angelesco ensemble converges to a new universallimiting process.In order to analyze the Angelesco ensembles we use the connection with multipleorthogonal polynomials. These polynomials are defined through orthogonalityrelations with respect to two or more weights, but they can also be characterizedby a matrix-valued Riemann-Hilbert (RH) problem. Subsequently we apply the Deift/Zhou method, which provides a factorization of the original matrix function.Additionally we use the Deift/Zhou method to study some universal limiting processesthat can intuitively be thought of as the result of collisions between two simplerprocesses. The considered processes depend on a double scaling parameter. Lettingthis parameter tend to infinity, combined with an appropriate scaling on the positionvariables, untangles the collisions.
Source: Lirias
Author: Deschout, Klaas
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7029835917
