Minimal Deterministic Echo State Networks Outperform Random Reservoirs in Learning Chaotic Dynamics
Not on the map. Why?
Abstract (arXiv)
Machine learning (ML) is widely used to model chaotic systems. Among ML approaches, echo state networks (ESNs) have received considerable attention due to their simple construction and fast training. However, ESN performance is highly sensitive to hyperparameter choices and to its random initialization. In this work, we demonstrate that ESNs constructed using deterministic rules and simple topologies (MESNs) outperform standard ESNs in the task of chaotic attractor reconstruction. We use a dataset of more than 90 chaotic systems to benchmark 10 different minimal deterministic reservoir initializations. We find that MESNs obtain up to a 41% reduction in error compared to standard ESNs. Furthermore, we show that the MESNs are more robust, exhibiting less inter-run variation, and have the ability to reuse hyperparameters across different systems. Our results illustrate how structured simplicity in ESN design can outperform stochastic complexity in learning chaotic dynamics.
Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Builds on 30 of its 90 references that are in the map. First posted on arXiv 8 Jul 2025.
Builds on
30 of 90 references.
References · 30
- 2021Next Generation Reservoir Computingcited by 160
- 2021Chaos as an interpretable benchmark for forecasting and data-driven modellingcited by 88
- 2018Echo state networks are universalcited by 87
- 2019Forecasting Chaotic Systems with Very Low Connectivity Reservoir Computerscited by 43
- 2018Reservoir computing approaches for representation and classification of multivariate time seriescited by 38
- 2019Model-free prediction of spatiotemporal dynamical systems with recurrent neural networks: Role of network spectral radiuscited by 34
- 2022Using Machine Learning to Anticipate Tipping Points and Extrapolate to Post-Tipping Dynamics of Non-Stationary Dynamical Systemscited by 33
- 2022A Systematic Exploration of Reservoir Computing for Forecasting Complex Spatiotemporal Dynamicscited by 30
- 2024Zero-shot forecasting of chaotic systemscited by 28
- 2021Robust Optimization and Validation of Echo State Networks for learning chaotic dynamicscited by 27
- 2023Generative learning for nonlinear dynamicscited by 22
- 2020Echo State Networks trained by Tikhonov least squares are L2(μ) approximators of ergodic dynamical systemscited by 21
18 more
- 2021Model-free inference of unseen attractors: Reconstructing phase space features from a single noisy trajectory using reservoir computingcited by 21
- 2025True Zero-Shot Inference of Dynamical Systems Preserving Long-Term Statisticscited by 20
- 2021Learning strange attractors with reservoir systemscited by 17
- 2025Panda: A pretrained forecast model for chaotic dynamicscited by 13
- 2016Determination of the edge of criticality in echo state networks through Fisher information maximizationcited by 12
- 2022Learning unseen coexisting attractorscited by 12
- 2023Attractor reconstruction with reservoir computers: The effect of the reservoir's conditional Lyapunov exponents on faithful attractor reconstructioncited by 9
- 2022ReservoirComputing.jl: An Efficient and Modular Library for Reservoir Computing Modelscited by 8
- 2023Simple Cycle Reservoirs are Universalcited by 8
- 2024Optimal Recurrent Network Topologies for Dynamical Systems Reconstructioncited by 8
- 2024Chaotic attractor reconstruction using small reservoirs -- the influence of topologycited by 7
- 2021Learning Continuous Chaotic Attractors with a Reservoir Computercited by 5
- 2024Machine Learning for Predicting Chaotic Systemscited by 4
- 2021Composing Modeling and Simulation with Machine Learning in Juliacited by 3
- 2024Universality of Real Minimal Complexity Reservoircited by 3
- 2024Linear Simple Cycle Reservoirs at the edge of stability perform Fourier decomposition of the input driving signalscited by 3
- 2025Deterministic Reservoir Computing for Chaotic Time Series Predictioncited by 3
- 2021Learning Hamiltonian dynamics by reservoir computercited by 2
Built on by
Citing papers per year: 2026: 2.
2 papers.