Rigorous Justification of the Fokker-Planck Equations of Neural Networks Based on an Iteration Perspective

Jian-Guo Liu,Ziheng Wang,Yuan Zhang,Zhennan Zhou
DOI: https://doi.org/10.1137/20m1338368
IF: 2.071
2022-01-01
SIAM Journal on Mathematical Analysis
Abstract:In this work, the primary goal is to establish rigorous connection betweenthe Fokker-Planck equation of neural networks with its microscopic model: thediffusion-jump stochastic process that captures the mean field behavior ofcollections of neurons in the integrate-and-fire model. The proof is based on anovel iteration scheme: with an auxiliary random variable counting the firingevents, both the density function of the stochastic process and the solution ofthe PDE problem admit series representations, and thus the difficulty inverifying the link between the density function and the PDE solution in eachsub problem has been greatly mitigated. The iteration approach provides ageneric frame in integrating the probability approach with PDE techniques, withwhich we prove that the density function of the diffusion-jump stochasticprocess is indeed the classical solution of the Fokker-Planck equation with aunique flux-shift structure.
What problem does this paper attempt to address?