A structure preserving numerical scheme for Fokker-Planck equations of structured neural networks with learning rules

DOI10.1137/21M1445600MaRDI QIDQ5097595zbMATH OpenOpenAlexWikidataFDO

Authors Qing He, Jingwei Hu, Zhennan Zhou

Publication date 25 August 2022

Published in SIAM Journal on Scientific Computing (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/2109.04667

zbMATH Keywords

Fokker-Planck equationmultiscale modelasymptotic preserving schemeneuron networkstructure preserving schemenonlinear noisy leaky integrate-and-fire model

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Neural networks for/in biological studies, artificial life and related topics (92B20) Asymptotic behavior of solutions to PDEs (35B40) Positive solutions to PDEs (35B09) Fokker-Planck equations (35Q84) PDEs in connection with biology, chemistry and other natural sciences (35Q92) PDEs with randomness, stochastic partial differential equations (35R60) Finite volume methods for initial value and initial-boundary value problems involving PDEs (65M08) Initial-boundary value problems for PDEs of mixed type (35M13)

Abstract: In this work, we are concerned with a Fokker-Planck equation related to the nonlinear noisy leaky integrate-and-fire model for biological neural networks which are structured by the synaptic weights and equipped with the Hebbian learning rule. The equation contains a small parameter $varepsilon$ separating the time scales of learning and reacting behavior of the neural system, and an asymptotic limit model can be derived by letting $varepsilono0$, where the microscopic quasi-static states and the macroscopic evolution equation are coupled through the total firing rate. To handle the endowed flux-shift structure and the multi-scale dynamics in a unified framework, we propose a numerical scheme for this equation that is mass conservative, unconditionally positivity preserving, and asymptotic preserving. We provide extensive numerical tests to verify the schemes' properties and carry out a set of numerical experiments to investigate the model's learning ability, and explore the solution's behavior when the neural network is excitatory.

Recommendations

• A structure preserving numerical scheme for Fokker-Planck equations of neuron networks: numerical analysis and exploration
• A numerical solver for a nonlinear Fokker-Planck equation representation of neuronal network dynamics
• Qualitative properties of solutions for the noisy integrate and fire model in computational neuroscience
• Structure preserving schemes for nonlinear Fokker-Planck equations and applications
• Rigorous justification of the Fokker-Planck equations of neural networks based on an iteration perspective

Cites work

• scientific article; zbMATH DE number 914383 (Why is no real title available?)
• A class of asymptotic-preserving schemes for the Fokker-Planck-Landau equation
• A finite volume scheme for nonlinear degenerate parabolic equations
• A finite-volume method for nonlinear nonlocal equations with a gradient flow structure
• A fully discrete positivity-preserving and energy-dissipative finite difference scheme for Poisson-Nernst-Planck equations
• A practical difference scheme for Fokker-Planck equations
• A second-order asymptotic-preserving and positivity-preserving exponential Runge-Kutta method for a class of stiff kinetic equations
• A structure preserving numerical scheme for Fokker-Planck equations of neuron networks: numerical analysis and exploration
• A unified structure preserving scheme for a multispecies model with a gradient flow structure and nonlocal interactions via singular kernels
• Analysis of nonlinear noisy integrate \& fire neuron models: blow-up and steady states
• Classical Solutions for a Nonlinear Fokker-Planck Equation Arising in Computational Neuroscience
• Distributed synaptic weights in a LIF neural network and learning rules
• Dynamics of sparsely connected networks of excitatory and inhibitory spiking neurons.
• Fully discrete positivity-preserving and energy-dissipating schemes for aggregation-diffusion equations with a gradient-flow structure
• Global solvability of a networked integrate-and-fire model of McKean-Vlasov type
• Global-in-time solutions and qualitative properties for the NNLIF neuron model with synaptic delay
• High order semi-implicit schemes for time dependent partial differential equations
• M-matrix characterizations. I: nonsingular M-matrices
• Particle systems with a singular mean-field self-excitation. Application to neuronal networks
• Positivity-preserving and asymptotic preserving method for 2D Keller-Segal equations
• Rigorous justification of the Fokker-Planck equations of neural networks based on an iteration perspective
• Structure preserving schemes for nonlinear Fokker-Planck equations and applications
• The mean-field equation of a leaky integrate-and-fire neural network: measure solutions and steady states

Cited in

(5)

• Rigorous justification of the Fokker-Planck equations of neural networks based on an iteration perspective
• A spectral method for a Fokker-Planck equation in neuroscience with applications in neural networks with learning rules
• A structure preserving numerical scheme for Fokker-Planck equations of neuron networks: numerical analysis and exploration
• Self-similar characteristics of neural networks based on Fokker-Planck equation
• A numerical solver for a nonlinear Fokker-Planck equation representation of neuronal network dynamics

This page was built for publication: A structure preserving numerical scheme for Fokker-Planck equations of structured neural networks with learning rules

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5097595)