Next Generation Reservoir Computing

Abstract

Reservoir computing is a best-in-class machine learning algorithm for processing information generated by dynamical systems using observed time-series data. Importantly, it requires very small training data sets, uses linear optimization, and thus requires minimal computing resources. However, the algorithm uses randomly sampled matrices to define the underlying recurrent neural network and has a multitude of metaparameters that must be optimized. Recent results demonstrate the equivalence of reservoir computing to nonlinear vector autoregression, which requires no random matrices, fewer metaparameters, and provides interpretable results. Here, we demonstrate that nonlinear vector autoregression excels at reservoir computing benchmark tasks and requires even shorter training data sets and training time, heralding the next generation of reservoir computing.

arXiv commentNew supplementary materials on noise sensitivity, dependence on training data set size and other smaller changes

References in corpus (13)

• Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
• Using Machine Learning to Replicate Chaotic Attractors and Calculate Lyapunov Exponents from Data
• Attractor Reconstruction by Machine Learning
• On Explaining the Surprising Success of Reservoir Computing Forecaster of Chaos? The Universal Machine Learning Dynamical System with Contrasts to VAR and DMD
• Forecasting Chaotic Systems with Very Low Connectivity Reservoir Computers
• Determination of the edge of criticality in echo state networks through Fisher information maximization
• Combining Machine Learning with Knowledge-Based Modeling for Scalable Forecasting and Subgrid-Scale Closure of Large, Complex, Spatiotemporal Systems
• Echo State Networks trained by Tikhonov least squares are L2(μ) approximators of ergodic dynamical systems
• How Entropic Regression Beats the Outliers Problem in Nonlinear System Identification
• Using Data Assimilation to Train a Hybrid Forecast System that Combines Machine-Learning and Knowledge-Based Components
• Robust Forecasting using Predictive Generalized Synchronization in Reservoir Computing
• Accuracy of neural networks for the simulation of chaotic dynamics: precision of training data vs precision of the algorithm
• Finding nonlinear system equations and complex network structures from data: a sparse optimization approach

Cited by in corpus (65)

• Time Series Quantum Reservoir Computing with Weak and Projective Measurements
• Model-free tracking control of complex dynamical trajectories with machine learning
• A Survey on Reservoir Computing and its Interdisciplinary Applications Beyond Traditional Machine Learning
• Perspectives on adaptive dynamical systems
• Combining machine learning and data assimilation to forecast dynamical systems from noisy partial observations
• Learning Spatiotemporal Chaos Using Next-Generation Reservoir Computing
• Model-free prediction of multistability using echo state network
• Optimizing Memory in Reservoir Computers
• Reservoir Computing and Photoelectrochemical Sensors: A Marriage of Convenience
• Parsimonious Physics-Informed Random Projection Neural Networks for Initial-Value Problems of ODEs and index-1 DAEs
• Learning unseen coexisting attractors
• Learning strange attractors with reservoir systems
• Controlling Chaos Using Edge Computing Hardware
• Catch-22s of reservoir computing
• Microwave signal processing using an analog quantum reservoir computer
• Forecasting trends in food security with real time data
• Model selection of chaotic systems from data with hidden variables using sparse data assimilation
• Time Shifts to Reduce the Size of Reservoir Computers
• Dynamical robustness of network of oscillators
• Controlling Chaotic Maps using Next-Generation Reservoir Computing
• Hybridizing Traditional and Next-Generation Reservoir Computing to Accurately and Efficiently Forecast Dynamical Systems
• Chaotic attractor reconstruction using small reservoirs -- the influence of topology
• Harnessing Quantum Extreme Learning Machines for image classification
• Symbolic Regression via Neural Networks
• Nonequilibrium Statistical Mechanics and Optimal Prediction of Partially-Observed Complex Systems
• Reservoir computing based on solitary-like waves dynamics of film flows: a proof of concept
• Symmetry-Aware Reservoir Computing
• Short-wavelength Reverberant Wave Systems for Physical Realization of Reservoir Computing
• Reservoir Computers with Random and Optimized Time-Shifts
• Neural Echo State Network using oscillations of gas bubbles in water
• Data-driven methods to estimate the committor function in conceptual ocean models
• Quantum Next Generation Reservoir Computing: An Efficient Quantum Algorithm for Forecasting Quantum Dynamics
• Practical Few-Atom Quantum Reservoir Computing
• Hyperparameter Optimization for Randomized Algorithms: A Case Study on Random Features
• Dynamical analysis of a parameter-aware reservoir computer
• Recurrent Spectral Network (RSN): shaping the basin of attraction of a discrete map to reach automated classification
• Quantum reservoir computing on random regular graphs
• Quantum Next-Generation Reservoir Computing and Its Quantum Optical Implementation
• Estimating the master stability function from the time series of one oscillator via reservoir computing
• Spintronic Physical Reservoir for Autonomous Prediction and Long-Term Household Energy Load Forecasting
• How more data can hurt: Instability and regularization in next-generation reservoir computing
• Budget-constrained Collaborative Renewable Energy Forecasting Market
• Temporal Convolution Derived Multi-Layered Reservoir Computing
• Denoising and Reconstruction of Nonlinear Dynamics using Truncated Reservoir Computing
• Parallel BiLSTM-Transformer networks for forecasting chaotic dynamics
• Exploring Non-Steady-State Charge Transport Dynamics in Information Processing: Insights from Reservoir Computing
• Practical Trainable Temporal Postprocessor for Multistate Quantum Measurement
• Reservoir Network with Structural Plasticity for Human Activity Recognition
• Data-Driven Forecasting of Non-Equilibrium Solid-State Dynamics
• Learning Beyond Experience: Generalizing to Unseen State Space with Reservoir Computing
• Tailored minimal reservoir computing: on the bidirectional connection between nonlinearities in the reservoir and in data
• Fast gradient-free activation maximization for neurons in spiking neural networks
• Next-Generation Reservoir Computing for Dynamical Inference
• Superconducting Qubit Readout Using Next-Generation Reservoir Computing
• From reductionism to realism: Holistic mathematical modelling for complex biological systems
• Enhanced Predictive Capability for Chaotic Dynamics by Modified Quantum Reservoir Computing
• Generating Quantum Reservoir State Representations with Random Matrices
• Adapting reservoir computing to solve the Schrödinger equation
• Phase transitions from linear to nonlinear information processing in neural networks
• Model-free inference of unseen attractors: Reconstructing phase space features from a single noisy trajectory using reservoir computing
• New analytic formulae for memory and prediction functions in reservoir computers with time delays
• Circuit design in biology and machine learning. I. Random networks and dimensional reduction
• Optimizing the Network Topology of a Linear Reservoir Computer
• Bayesian Optimization of Multi-Bit Pulse Encoding in In2O3/Al2O3 Thin-film Transistors for Temporal Data Processing
• Learning stochastic filtering