On the structure of quasi-stationary competing particle systems

Ordinary differential equations and systems with randomness (34F05) Point processes (e.g., Poisson, Cox, Hawkes processes) (60G55) Stochastic ordinary differential equations (aspects of stochastic analysis) (60H10)

Abstract: We study point processes on the real line whose configurations

X

are locally finite, have a maximum and evolve through increments which are functions of correlated Gaussian variables. The correlations are intrinsic to the points and quantified by a matrix

Q = {q_{i j}}_{i, j i n m a t h b b N}

. A probability measure on the pair

(X, Q)

is said to be quasi-stationary if the joint law of the gaps of

X

and of

Q

is invariant under the evolution. A known class of universally quasi-stationary processes is given by the Ruelle Probability Cascades (RPC), which are based on hierarchically nested Poisson--Dirichlet processes. It was conjectured that up to some natural superpositions these processes exhausted the class of laws which are robustly quasi-stationary. The main result of this work is a proof of this conjecture for the case where

q_{i j}

assume only a finite number of values. The result is of relevance for mean-field spin glass models, where the evolution corresponds to the cavity dynamics, and where the hierarchical organization of the Gibbs measure was first proposed as an ansatz.

Review text

A competing particle system is a point process on \({\mathbb R}\) whose configurations \(X\) is given by: 1. an ordered sequence of positions, \(X=\{X_i\}_{i\in {\mathbb N}}\subset {\mathbb R}\), with \(X_1\geq X_2\geq \dots\), and 2. a so-called overlap matrix, which is a positive definite quadratic form given by \(Q=\{q_{ij}\}\), with \(q_{ii}=1\) for all \(i\in {\mathbb N}\) (and hence \(|q_{ij}|\leq 1\)). The time evolution is given by correlated increments, \(X_i\mapsto X_i+\Psi(k_i)\), where \(k\) is a Gaussian field on \({\mathbb N}\), independent of \(X\) with covariance given by the matrix \(Q\), and \(\Psi\) is a real function. The paper considers the question of describing the collection of distributions of \((X, Q)\) which are quasi-stationary under the above evolution, in the sense that the joint law of the gaps, \(\{X_i - X_{i+1}\}_{i\in{\mathbb N}}\), and \(q_{ij}\) is invariant under the evolution. In the particular case where \(Q\) is the identity matrix, that is the increments are independent, a quasi-stationary process \(X\) must be a superposition of Poisson processes with exponential density, see [\textit{A. Ruzmaikina} and \textit{M. Aizenman}, Ann. Probab. 33, No. 1, 82--113 (2005; Zbl 1096.60042)]. The Ruelle probability cascades form a broader class of processes, with \(Q\) not limited to the identity matrix. Each process in this collection is quasi-stationary under any dynamics of the above form with an exponentially bounded \(\Psi\), and the class is parameterized by probability measures on \([0,1]\). It was conjectured that this collection includes all processes which are robustly quasi-stationary. The significance of this conjecture for spin glass models is discussed in [\textit{M. Aizenman, R. Sims} and \textit{S. L. Starr}, in: Prospects in mathematical physics. Young researchers symposium of the 14th international congress on mathematical physics, Lisbon, Portugal, July 25--26, 2003. Providence, RI: American Mathematical Society (AMS). Contemporary Mathematics 435, 1--30 (2007; Zbl 1175.82033)]. The main result of the paper establishes the above conjecture in the case of systems that are quasi-stationary in a robust sense and for which the set of values taken by the covariances of the point \(i\) is finite.

Recommendations

Cites work

Cited in

(49)

This page was built for publication: On the structure of quasi-stationary competing particle systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2270610)