Estimating the master stability function from the time series of one oscillator via reservoir computing
arXiv:2304.13125 · doi:10.1103/PhysRevE.108.L032201
Abstract
The master stability function (MSF) yields the stability of the globally synchronized state of a network of identical oscillators in terms of the eigenvalues of the adjacency matrix. In order to compute the MSF, one must have an accurate model of an uncoupled oscillator, but often such a model does not exist. We present a reservoir computing technique for estimating the MSF given only the time series of a single, uncoupled oscillator. We demonstrate the generality of our technique by considering a variety of coupling configurations of networks consisting of Lorenz oscillators or H{é}non maps.
References in corpus (16)
- Synchronization in small-world systems
- Spontaneous synchrony in power-grid networks
- Using Machine Learning to Replicate Chaotic Attractors and Calculate Lyapunov Exponents from Data
- Symmetries, Cluster Synchronization, and Isolated Desynchronization in Complex Networks
- Next Generation Reservoir Computing
- Attractor Reconstruction by Machine Learning
- Coupling functions: Universal insights into dynamical interaction mechanisms
- The synchronized dynamics of time-varying networks
- Adaptive synchronization of dynamics on evolving complex networks
- Synchronizing distant nodes: a universal classification of networks
- Robustness of Optimal Synchronization in Real Networks
- Anticipating synchronization with machine learning
- Topological Control of Synchronization Patterns: Trading Symmetry for Stability
- Parallel Machine Learning for Forecasting the Dynamics of Complex Networks
- Using Machine Learning to Assess Short Term Causal Dependence and Infer Network Links
- Machine Learning Link Inference of Noisy Delay-coupled Networks with Opto-Electronic Experimental Tests