Random and optimal configurations in complex function theory

Abstract: This thesis consists of six articles spanning over several areas of mathematical analysis. The dominant theme is the study of random point processes and optimal point configurations, which may be though of as systems of charged particles with mutual repulsion. We are predominantly occupied with questions of universality, a phenomenon that appears in the study of random complex systems where seemingly unrelated microscopic laws produce systems with striking similarities in various scaling limits. In particular, we obtain a complete asymptotic expansion of planar orthogonal polynomials with respect to exponentially varying weights, which yields universality for the microscopic boundary behavior in the random normal matrix (RNM) model (Paper A) as well as in the case of more general interfaces for Bergman kernels (Paper B). Still in the setting of RNM ensembles, we investigate properties of scaling limits near singular points of the boundary of the spectrum, including cusps points (Paper C). We also obtain a central limit theorem for fluctuations of linear statistics in the polyanalytic Ginibre ensemble, using a new representation of the polyanalytic correlation kernel in terms of algebraic differential operators acting on the classical Ginibre kernel (Paper D). Paper E is concerned with an extremal problem for analytic polynomials, which may heuristically be interpreted as an optimal packing problem for the corresponding zeros. The last article (Paper F) concerns a different theme, namely a sharp topological transition in an Lp-analogue of classical Carleman classes for 0 < p < 1.