Using reservoir computer to predict and prevent extreme events

Abstract

We show that a reservoir computer is an effective tool for model-free prediction of extreme events in deterministic chaotic systems. This prediction allows us to suppress unwanted extreme events, by applying weak control perturbations to the system at times preceding expected extreme events. The effectiveness of such a prediction and prevention strategy is demonstrated for a system of globally coupled FitzHugh-Nagumo neurons and for a system of two almost identical unidirectionally coupled chaotic oscillators.

References in corpus (13)

• A model for cascading failures in complex networks
• Using Machine Learning to Replicate Chaotic Attractors and Calculate Lyapunov Exponents from Data
• Attractor Reconstruction by Machine Learning
• Hybrid Forecasting of Chaotic Processes: Using Machine Learning in Conjunction with a Knowledge-Based Model
• Dragon-kings: mechanisms, statistical methods and empirical evidence
• Predictability and suppression of extreme events in complex systems
• Reservoir computing with a single time-delay autonomous Boolean node
• Extreme events in excitable systems and mechanisms of their generation
• Route to extreme events in excitable systems
• Rapid Time Series Prediction with a Hardware-Based Reservoir Computer
• Machine learning algorithms for predicting the amplitude of chaotic laser pulses
• A primer on noise-induced transitions in applied dynamical systems
• Dynamical indicators for the prediction of bursting phenomena in high-dimensional systems

Cited by in corpus (7)

• Extreme events in dynamical systems and random walkers: A review
• Extreme events in globally coupled chaotic maps
• Optimized ensemble deep learning framework for scalable forecasting of dynamics containing extreme events
• Dynamical system analysis of a data-driven model constructed by reservoir computing
• Model-free prediction of emergence of extreme events in a parametrically driven nonlinear dynamical system by Deep Learning
• Model-assisted deep learning of rare extreme events from partial observations
• Choose your tools carefully: A Comparative Evaluation of Deterministic vs. Stochastic and Binary vs. Analog Neuron models for Implementing Emerging Computing Paradigms