Central limit theorems for global and local empirical measures of diffusions on Erdős-Rényi graphs (Q6177575)

From MaRDI portal
scientific article; zbMATH DE number 7790322
Language Label Description Also known as
English
Central limit theorems for global and local empirical measures of diffusions on Erdős-Rényi graphs
scientific article; zbMATH DE number 7790322

    Statements

    Central limit theorems for global and local empirical measures of diffusions on Erdős-Rényi graphs (English)
    0 references
    0 references
    0 references
    0 references
    17 January 2024
    0 references
    The authors consider fluctuations of the local and global empirical measures of diffusions interacting on an asymmetric Erdős-Rényi graph on \(n\) vertices. This graph is defined via the collection \((\xi^{(n)}_{ij})_{i,j=1,\ldots,n}\) of independent Bernoulli random variables, each with mean \(p_n\). Each of the \(n\) particles has motion represented by a function \(\theta_t^{i,n}\) on the torus \(\mathbb{T}=\mathbb{R}/2\pi\mathbb{Z}\), with the corresponding dynamics defined via the equations \[ \text{d}\theta_t^{i,n}=\frac{1}{np_n}\sum_{j=1}^n\xi_{ij}^{(n)}\Gamma(\theta_t^{i,n},\theta_t^{j,n})\text{d}t+\text{d}B_t^i \] for \(i=1,\ldots,n\), with initial conditions \((\theta_0^{i,n})_{i=1,\ldots,n}\). Here the \(B^i\) are independent standard Brownian motions, also independent of the initial conditions, and \(\Gamma:\mathbb{T}\times\mathbb{T}\to\mathbb{R}\) is a regular function. The main results of the present paper are limits for the standard fluctuations of the global empirical measure \[ \frac{1}{n}\sum_{i=1}^n\delta_{\theta_t^{i,n}} \] around its limit as \(n\to\infty\), and similarly for the local empirical measure \[ \frac{1}{np_n}\sum_{i=1}^n\xi^{(n)}_{li}\delta_{\theta_t^{i,n}}\,, \] of particles at unit distance from vertex \(l\) in the graph. The proofs employ generalised Grothendieck inequalities which allow the decorrelation of the dynamics from measures of proximity of the underlying Erdős-Rényi graph to the complete graph.
    0 references
    0 references
    central limit theorem
    0 references
    empirical measures
    0 references
    Erdős-Rényi random graphs
    0 references
    Grothendieck inequalities
    0 references
    interacting diffusions
    0 references
    mean-field systems
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references