A dynamic learning method for phase synchronization control in Hodgkin–Huxley neuronal networks

DOI: https://doi.org/10.1140/epjs/s11734-024-01171-w

2024-04-24

The European Physical Journal Special Topics

Abstract:The mathematical optimization techniques may control the network to target firing patterns by adjusting the weights of network nodes. Inspired by the dynamics of dynamical learning, we recently proposed a technique for dynamic learning of synchronous (DLS) to control the firing state of neural networks. In this study, we apply the DLS technique to a Hodgkin–Huxley-style neural network, and investigate in regular, random, small-world and scale-free networks. We use the DLS technique to accomplish the network adaptive global synchronization, adaptive local synchronization, and phase locking with a single supervisory node. Furthermore, we investigated the robustness of the DLS technique in noisy environments and find that the DLS technique demonstrates remarkable effectiveness even in the presence of weak noise. However, in scenarios with stronger noise, there is a trade-off between optimizing training and avoiding overfitting, i.e., a too narrow weight adjustment range may hinder training effectiveness, while an excessively wide range results in abnormal node firing dynamics. We expect the DLS technique to be potentially valuable for more studies of nonlinear systems.

physics, multidisciplinary