[Paper Review] Generalized $\beta$-Gaussian Ensemble Equilibrium measure method

Mohamed Bouali|arXiv (Cornell University)|Aug 30, 2014

Random Matrices and Applications6 references3 citations

TL;DR

This paper generalizes the $eta$-Gaussian ensemble by introducing a parameter $eta > 0$ and an additional $|x_i|^{2 heta}$ weight, using logarithmic potential theory and equilibrium measure methods to prove that the eigenvalue density converges weakly to a limiting probability measure as $n \to \infty$. The key result is a generalized Wigner semi-circle law for arbitrary $\beta > 0$, extending classical results for $\beta = 1,2,4$.

ABSTRACT

We investigate $\\beta$-Generalized random Hermitian matrices ensemble sometimes called Chiral ensemble. We give global asymptotic of the density of eigenvalues or the statistical density. We investigate general method names as equilibrium measure method. When taking $n$ large limit we will see that the asymptotic density of eigenvalues generalize the Wigner semi-circle law.

Motivation & Objective

• To extend the classical Wigner semi-circle law to generalized $\beta$-Gaussian ensembles with $\beta > 0$ and additional $|x|^{2\theta}$ weights.
• To establish the weak convergence of the empirical eigenvalue measure to a limiting equilibrium measure via potential-theoretic methods.
• To compute the exact energy of the limiting equilibrium measure for general $\beta > 0$, generalizing known results for $\beta = 2$.
• To provide a rigorous asymptotic analysis of eigenvalue density using logarithmic potential theory and compactness arguments in the weak topology.

Proposed method

• Uses the joint probability density function of eigenvalues with a $\beta$-Hermite-type interaction and a power-law weight $|x_i|^{2\theta}$.
• Applies logarithmic potential theory to define the energy functional and equilibrium measure associated with the potential $Q_{\alpha_n}(x) = x^2 + 2\alpha_n \log(1/|x|)$.
• Employs the method of steepest descent via the function $K_n(x)$, which captures the energy of the system and concentrates the measure near its infimum.
• Uses compactness and weak convergence arguments to show that the empirical eigenvalue measure $\nu_n$ converges to a limiting probability measure $\nu_{\beta,c}$.
• Relies on boundary value distributions of holomorphic functions and the Cauchy transform to characterize the equilibrium measure and its properties.
• Proves concentration of measure via the set $A_{n,\varepsilon}$, showing $\mathbb{P}_{n,\mu_n}(A_{n,\varepsilon}) \to 1$ as $n \to \infty$.

Experimental results

Research questions

• RQ1 How does the eigenvalue density of the generalized $\beta$-Gaussian ensemble behave in the large $n$ limit when $\beta > 0$ and a power-law weight $|x|^{2\theta}$ is introduced?
• RQ2 Can the Wigner semi-circle law be generalized to arbitrary $\beta > 0$ using equilibrium measure theory?
• RQ3 What is the exact value of the energy of the limiting equilibrium measure for general $\beta > 0$, and how does it compare to the $\beta = 2$ case?
• RQ4 How does the inclusion of the $|x|^{2\theta}$ weight affect the asymptotic eigenvalue distribution and the associated potential theory?

Key findings

• The empirical eigenvalue measure $\nu_n$ converges weakly to a limiting probability measure $\nu_{\beta,c}$ as $n \to \infty$, under appropriate rescaling by $\sqrt{n}$.
• The limiting measure $\nu_{\beta,c}$ is an equilibrium measure for the potential $Q_{\alpha_n}(x) = x^2 + 2\alpha_n \log(1/|x|)$ with $\alpha_n = \mu_n / n$.
• The energy of the limiting measure is $E^*_{\beta,c}$, and the paper computes this energy for general $\beta > 0$, extending known results for $\beta = 2$.
• The measure $\nu_n$ concentrates in a neighborhood of the set $A_{n,\varepsilon}$ where the energy functional $K_n(x)$ is minimized, and $\mathbb{P}_{n,\mu_n}(A_{n,\varepsilon}) \to 1$ as $n \to \infty$.
• The limiting measure $\nu_{\beta,c}$ generalizes the Wigner semi-circle law, reducing to it in the $\beta = 2$, $\theta = 0$ case.
• The convergence is established via weak compactness and boundary value analysis of holomorphic functions, particularly using the Cauchy transform and logarithmic potentials.