[Paper Review] Interpretations of some parameter dependent generalizations of classical matrix ensembles

Peter J. Forrester, Eric M. Rains|ArXiv.org|Nov 19, 2002

Random Matrices and Applications29 references4 citations

TL;DR

This paper provides probabilistic interpretations of parameter-dependent generalizations of classical matrix ensembles—specifically, those interpolating between orthogonal, unitary, and symplectic symmetries—via the continuous Robinson-Schensted-Knuth correspondence and non-intersecting lattice paths. It establishes a link to Anderson’s Selberg integral work and constructs random matrix models whose eigenvalue PDFs realize the original parameter-dependent distributions, enabling exact sampling via zeros of random three-term recurrences.

ABSTRACT

Two types of parameter dependent generalizations of classical matrix ensembles are defined by their probability density functions (PDFs). As the parameter is varied, one interpolates between the eigenvalue PDF for the superposition of two classical ensembles with orthogonal symmetry and the eigenvalue PDF for a single classical ensemble with unitary symmetry, while the other interpolates between a classical ensemble with orthogonal symmetry and a classical ensemble with symplectic symmetry. We give interpretations of these PDFs in terms of probabilities associated to the continuous Robinson-Schensted-Knuth correspondence between matrices, with entries chosen from certain exponential distributions, and non-intersecting lattice paths, and in the course of this probability measures on partitions and pairs of partitions are identified. The latter are generalized by using Macdonald polynomial theory, and a particular continuum limit -- the Jacobi limit -- of the resulting measures is shown to give PDFs related to those appearing in the work of Anderson on the Selberg integral. By interpreting Anderson's work as giving the PDF for the zeros of a certain rational function, it is then possible to identify random matrices whose eigenvalue PDFs realize the original parameter dependent PDFs. This line of theory allows sampling of the original parameter dependent PDFs, their Anderson-type generalizations and associated marginal distributions, from the zeros of certain polynomials defined in terms of random three term recurrences.

Motivation & Objective

• To interpret two families of parameter-dependent matrix ensemble PDFs that interpolate between classical ensembles with orthogonal, unitary, and symplectic symmetry.
• To connect these PDFs to probabilistic models involving matrices with exponential entries and non-intersecting lattice paths.
• To generalize partition and pair-partition measures using Macdonald polynomial theory.
• To identify a Jacobi limit of these measures that recovers PDFs related to Anderson’s Selberg integral work.
• To construct random matrix models whose eigenvalue distributions match the original parameter-dependent PDFs, enabling exact sampling via polynomial zeros.

Proposed method

• Uses the continuous Robinson-Schensted-Knuth (RSK) correspondence to map random matrices with i.i.d. exponential entries to non-intersecting lattice paths.
• Applies Macdonald polynomial theory to generalize measures on partitions and pairs of partitions arising from the RSK correspondence.
• Derives a Jacobi limit of these generalized measures, showing convergence to PDFs related to Anderson’s Selberg integral results.
• Constructs a random matrix model via a sum of a fixed diagonal matrix, a Wishart-type matrix, and a GUE matrix, with eigenvalue PDF derived via Laplace transforms and the Harish-Chandra/Itzykson-Zuber formula.
• Uses the inverse Laplace transform of generating functions to express the eigenvalue PDF of the constructed random matrix.
• Demonstrates that eigenvalue sampling is possible by computing the zeros of random three-term recurrence polynomials derived from the matrix model.

Experimental results

Research questions

• RQ1 How can parameter-dependent generalizations of classical matrix ensembles be interpreted probabilistically through combinatorial structures like the RSK correspondence and non-intersecting lattice paths?
• RQ2 What is the connection between these generalized measures and Anderson’s work on the Selberg integral and its associated eigenvalue PDFs?
• RQ3 Can random matrix models be explicitly constructed such that their eigenvalue PDFs match the original parameter-dependent PDFs?
• RQ4 How does Macdonald polynomial theory extend the measures on partitions and pairs of partitions derived from the RSK correspondence?
• RQ5 What is the role of the Jacobi limit in relating the generalized measures to known classical ensembles?

Key findings

• The parameter-dependent PDFs (1.1) and (1.4) interpolate between orthogonal and symplectic ensembles, while (1.2) and (1.5) interpolate between superimposed orthogonal and unitary ensembles.
• The PDFs are interpreted via the continuous RSK correspondence between matrices with exponential entries and non-intersecting lattice paths, yielding measures on partitions and pairs of partitions.
• Macdonald polynomial theory generalizes these partition measures, and their Jacobi limit recovers PDFs related to Anderson’s Selberg integral work.
• A random matrix model is constructed as $ M = UAU^ abla + XBX^ abla + abla{t}Y $, whose eigenvalue PDF is explicitly derived using Laplace transforms and the Harish-Chandra/Itzykson-Zuber formula.
• The eigenvalue PDF of this model matches the original parameter-dependent PDFs, and sampling is achieved via the zeros of random three-term recurrence polynomials.
• In special cases—such as $ t=0 $, $ B = \text{diag}(b,0,\dots,0) $, or $ B=0 $—the model recovers known results, including the Laguerre and GUE eigenvalue distributions.