Abstract:For simulating large networks of neurons Hines proposed a method which uses extensively the structure of the arising systems of ordinary differential equations in order to obtain an efficient implementation. The original method requires constant step sizes and produces the solution on a staggered grid. In the present paper a one-step modification of this method is introduced and analyzed with respect to their stability properties. The new method allows for step size control. Local error estimators are constructed. The method has been implemented in matlab and tested using simple Hodgkin-Huxley type models. Comparisons with standard state-of-the-art solvers are provided.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to improve the Hines method to allow variable - step - size control while maintaining computational efficiency and accuracy. Specifically: 1. **Limitations of the original Hines method**: - The Hines method is very effective in simulating large - scale neural networks because it makes full use of the structure of the system of ordinary differential equations to achieve efficient numerical solutions. - However, the original Hines method requires the use of a fixed step size and generates solutions on a staggered grid, which limits its flexibility and adaptability. 2. **Introduction of variable - step - size modification**: - The paper proposes a new one - step method modification that allows variable - step - size control. This method not only retains the efficiency of the Hines method but also improves flexibility, enabling it to better adapt to the needs of different problems. - The new method achieves step - size control by constructing a local error estimator, thereby ensuring the accuracy of the solution. 3. **Stability analysis**: - The paper conducts a detailed analysis of the stability of the new method, especially for stiff differential equation systems. By introducing appropriate test equations, the asymptotic stability of the method is studied. - The results show that in many cases, these stability regions are bounded, which is crucial for choosing an appropriate step size. 4. **Numerical experiments and comparisons**: - The paper conducts numerical experiments through the simple Hodgkin - Huxley model and other standard solvers to verify the effectiveness and superiority of the new method. - The experimental results show that the new method performs well under low - tolerance conditions, and especially in cases where high precision is required, the variable - step - size method is competitive. In summary, this paper aims to improve the Hines method by introducing variable - step - size control, making it more flexible and efficient in simulating large - scale neural networks while maintaining the accuracy and stability of the numerical solution.

On a variable step size modification of Hines' method in computational neuroscience

Refactorization of a variable step, unconditionally stable method of Dahlquist, Liniger and Nevanlinna

Multiscale modeling via split-step methods in neural firing

The Seven-League Scheme: Deep Learning for Large Time Step Monte Carlo Simulations of Stochastic Differential Equations

On the convergence of adaptive approximations for stochastic differential equations

STEPS 4.0: Fast and memory-efficient molecular simulations of neurons at the nanoscale

Asymptotic Behaviour of Time Stepping Methods for Phase Field Models

A Self-Organizing State-Space-Model Approach for Parameter Estimation in Hodgkin-Huxley-Type Models of Single Neurons

Adaptive stepsize algorithms for Langevin dynamics

Efficient optimization of ODE neuron models using gradient descent

Laws of Large Numbers and Langevin Approximations for Stochastic Neural Field Equations

An Antithetic Multilevel Monte Carlo-Milstein Scheme for Stochastic Partial Differential Equations

Structure-preserving numerical integrators for Hodgkin-Huxley-type systems

System Level Numerical Analysis of a Monte Carlo Simulation of the E. Coli Chemotaxis

Heuristic Tree-Partition-Based Parallel Method for Biophysically Detailed Neuron Simulation

Distributed Inexact Newton Method with Adaptive Step Sizes

Atomically Detailed Simulations of Helix Formation with the Stochastic Difference Equation

Step Size Matters in Deep Learning

Self-oscillations in ring Toda chains with negative friction.

Multiscale Computations on Neural Networks: From the Individual Neuron Interactions to the Macroscopic-Level Analysis

Comparison Tests of Variable-Stepsize Algorithms for Stochastic Ordinary Differential Equations of Finance