Abstract:We consider two classes of natural stochastic processes on finite unlabeled graphs. These are Euclidean stochastic optimization algorithms on the adjacency matrix of weighted graphs and a modified version of the Metropolis MCMC algorithm on stochastic block models over unweighted graphs. In both cases we show that, as the size of the graph goes to infinity, the random trajectories of the stochastic processes converge to deterministic curves on the space of measure-valued graphons. Measure-valued graphons, introduced by Lovász and Szegedy in \cite{lovasz2010decorated}, are a refinement of the concept of graphons that can distinguish between two infinite exchangeable arrays that give rise to the same graphon limit. We introduce new metrics on this space which provide us with a natural notion of convergence for our limit theorems. This notion is equivalent to the convergence of infinite-exchangeable arrays. Under suitable assumptions and a specified time-scaling, the Metropolis chain admits a diffusion limit as the number of vertices go to infinity. We then demonstrate that, in an appropriately formulated zero-noise limit, the stochastic process of adjacency matrices of this diffusion converges to a deterministic gradient flow curve on the space of graphons introduced in\cite{Oh2023}. A novel feature of this approach is that it provides a precise exponential convergence rate for the Metropolis chain in a certain limiting regime. The connection between a natural Metropolis chain commonly used in exponential random graph models and gradient flows on graphons, to the best of our knowledge, is new in the literature as well.

Graphon-valued stochastic processes from population genetics

The grapheme-valued Wright-Fisher diffusion with mutation

Stochastic Graphon Mean Field Games with Jumps and Approximate Nash Equilibria

Continuum graph dynamics via population dynamics: well-posedness, duality and equilibria

On the stochastic evolution of finite populations

Stochastic Graphon Games with Memory

Path convergence of Markov chains on large graphs

Stochastic PDEs on graphs as scaling limits of discrete interacting systems

Convergence of genealogies through spinal decomposition with an application to population genetics

Large population limits of Markov processes on random networks

Stochastic partial differential equations describing neutral genetic diversity under short range and long range dispersal

Transition probabilities for degenerate diffusions arising in population genetics

Stochastic Graphon Games: I. The Static Case

Graphon Particle Systems, Part I: Spatio-Temporal Approximation and Law of Large Numbers

Nonequilibrium steady-state dynamics of Markov processes on graphs

Interacting stochastic processes on sparse random graphs

Evolutionary graph theory beyond single mutation dynamics: on how network structured populations cross fitness landscapes

Evolutionary games on isothermal graphs

Sampling probabilities, diffusions, ancestral graphs, and duality under strong selection

Interacting particle systems on sparse $W$-random graphs

Generalized Graphon Process: Convergence of Graph Frequencies in Stretched Cut Distance