A stochastic system with infinite interacting components to model the time evolution of the membrane potentials of a population of neurons (Q300614)

A new class of interacting particle systems with a countable number of components having natural state space and interactions of infinite range is introduced. The model describes the evolution of the membrane potentials of interacting neurons in time. The interactions between neurons is defined by directed graphs that can change according to the membrane potential. The author has been inspired by the models introduced in [\textit{A. Galves} and \textit{E. Löcherbach}, J. Stat. Phys. 151, No. 5, 896--921 (2013; Zbl 1276.82046)]. The second section is devoted to definitions and results. One introduces the model and the result that claims the existence and uniqueness of the defined stationary process. A second theorem refers to the existence of a perfect simulation algorithm based on a coupled construction of two processes defined in the previous result. It provides also an upper bound for the error while sampling from a finite set of neurons instead from the whole system. The perfect simulation is developed in the third section and the proof of theorems is done in the last sections.