Networks of Noisy Neurons
Time: 2025-03-26
Published By: Ruixin Li
Speaker(s): Ben Hambly (University of Oxford)
Time: 16:00-17:00 April 8, 2025
Venue: Room 77201, Jingchunyuan 78, BICMR
Abstract: We consider a network of integrate-and-fire models for neurons subject to common noise and analyse its mean field limit. The neurons are described by their membrane potential, which diffuses until a threshold is reached, when the neuron fires, that is it resets to a resting potential and transmits voltage to other neurons. We show that in the mean field limit, the evolution of the limiting membrane potential density can be described by a stochastic non-local non-linear Fokker-Planck equation. We investigate when there can be synchronization of the firing of non-trivial proportions of the network.
Bio-Sketch: Ben Hambly is Professor of Mathematics at University of Oxford. His research interests are in probability, stochastic processes, financial mathematics and fractals.