Large sampling intervals for learning and predicting chaotic systems with reservoir computing DOI

Qingyan Xie,

Zixiang Yan, Hui Zhao

и другие.

Journal of Physics A Mathematical and Theoretical, Год журнала: 2024, Номер 57(29), С. 295702 - 295702

Опубликована: Июль 9, 2024

Abstract Reservoir computing (RC) is an efficient artificial neural network for model-free prediction and analysis of dynamical systems time series. As a data-based method, the capacity RC strongly affected by sampling interval training data. In this paper, taking Lorenz system as example, we explore influence on performance in predicting chaotic sequences. When increases, first enhanced then weakened, presenting bell-shaped curve. By slightly revising calculation method output matrix, with small can be improved. Furthermore, learn reproduce state large interval, which almost five times larger than that classic fourth-order Runge–Kutta method. Our results show applications where intervals are constrained laid foundation building fast algorithm iteration steps.

Язык: Английский

Time‐Multiplexed Reservoir Computing with Quantum‐Dot Lasers: Impact of Charge‐Carrier Scattering Timescale DOI Creative Commons
Huifang Dong, Lina Jaurigue, Kathy Lüdge

и другие.

physica status solidi (RRL) - Rapid Research Letters, Год журнала: 2025, Номер unknown

Опубликована: Фев. 3, 2025

Reservoir computing with optical devices offers an energy‐efficient approach for time‐series forecasting. Quantum dot lasers feedback are modeled in this article to explore the extent which increased complexity charge‐carrier dynamics within nanostructured semiconductor can enhance prediction performance. By tuning scattering interactions, laser's and response time be finely adjusted, allowing a systematic investigation. It is found that both system task requirements need considered find optimal operation conditions. Further, pronounced relaxation oscillations outperform those strongly damped dynamics, even if underlying more complex. This demonstrates reservoir performance relies not only on high internal phase space dimension but also effective utilization of these through output sampling process, quantum laser, computing, delay, rate, oscillation.

Язык: Английский

Процитировано

0

Reservoir computing with state-dependent time delay DOI
G. O. Danilenko, Anton V. Kovalev, D. S. Citrin

и другие.

Physical review. E, Год журнала: 2025, Номер 111(3)

Опубликована: Март 27, 2025

We examine a new design of reservoir computing based on an otherwise linear dynamical system subject to feedback in which delay time linearly depends the system's state. Despite apparent linearity under casual perusal, nonetheless possesses nonlinearity that can be used for time-delay computing. find close multiple Hopf bifurcation points lead rich sawtooth-shaped transient response input signals, beneficial capabilities. benchmark memory capacity and performance solving delayed XOR, Iris flower classification tasks, Santa Fe time-series prediction task. demonstrate how tuned by changing dependence.

Язык: Английский

Процитировано

0

The Drosophila Connectome as a Computational Reservoir for Time-Series Prediction DOI Creative Commons
Leone Costi, Alexander Hadjiivanov, Dominik Dold

и другие.

Biomimetics, Год журнала: 2025, Номер 10(5), С. 341 - 341

Опубликована: Май 21, 2025

In this work, we explore the possibility of using topology and weight distribution connectome a Drosophila, or fruit fly, as reservoir for multivariate chaotic time-series prediction. Based on information taken from recently released full connectome, create connectivity matrix an Echo State Network. Then, use only most connected neurons implement two possible selection criteria, either preserving breaking relative proportion different neuron classes which are also included in documented to obtain computationally convenient reservoir. We then investigate performance such architectures compare them state-of-the-art reservoirs. The results show that connectome-based architecture is significantly more resilient overfitting compared standard implementation, particularly cases already prone overfitting. To further isolate role synaptic weights, hybrid reservoirs with but random weights topologies study, demonstrating both factors play increased resilience. Finally, perform experiment where entire used Despite much higher number trained parameters, remains has lower normalized error, under 2%, at regularisation, all other regularisation.

Язык: Английский

Процитировано

0

Large sampling intervals for learning and predicting chaotic systems with reservoir computing DOI

Qingyan Xie,

Zixiang Yan, Hui Zhao

и другие.

Journal of Physics A Mathematical and Theoretical, Год журнала: 2024, Номер 57(29), С. 295702 - 295702

Опубликована: Июль 9, 2024

Abstract Reservoir computing (RC) is an efficient artificial neural network for model-free prediction and analysis of dynamical systems time series. As a data-based method, the capacity RC strongly affected by sampling interval training data. In this paper, taking Lorenz system as example, we explore influence on performance in predicting chaotic sequences. When increases, first enhanced then weakened, presenting bell-shaped curve. By slightly revising calculation method output matrix, with small can be improved. Furthermore, learn reproduce state large interval, which almost five times larger than that classic fourth-order Runge–Kutta method. Our results show applications where intervals are constrained laid foundation building fast algorithm iteration steps.

Язык: Английский

Процитировано

1