Large-Scale Optical Reservoir Computing for Spatiotemporal Chaotic Systems Prediction

Abstract

Reservoir computing is a relatively recent computational paradigm that originates from a recurrent neural network and is known for its wide range of implementations using different physical technologies. Large reservoirs are very hard to obtain in conventional computers, as both the computation complexity and memory usage grow quadratically. We propose an optical scheme performing reservoir computing over very large networks potentially being able to host several millions of fully connected photonic nodes thanks to its intrinsic properties of parallelism and scalability. Our experimental studies confirm that, in contrast to conventional computers, the computation time of our optical scheme is only linearly dependent on the number of photonic nodes of the network, which is due to electronic overheads, while the optical part of computation remains fully parallel and independent of the reservoir size. To demonstrate the scalability of our optical scheme, we perform for the first time predictions on large spatiotemporal chaotic datasets obtained from the Kuramoto-Sivashinsky equation using optical reservoirs with up to 50 000 optical nodes. Our results are extremely challenging for conventional von Neumann machines, and they significantly advance the state of the art of unconventional reservoir computing approaches, in general.

arXiv comment9 pages, 7 figures, abstract is modified according to the published version

References in corpus (24)

• Recent Advances in Physical Reservoir Computing: A Review
• Measuring the Transmission Matrix in Optics : An Approach to the Study and Control of Light Propagation in Disordered Media
• Optoelectronic Reservoir Computing
• Light fields in complex media: mesoscopic scattering meets wave control
• Using Machine Learning to Replicate Chaotic Attractors and Calculate Lyapunov Exponents from Data
• All Optical Neural Network with Nonlinear Activation Functions
• Reinforcement Learning in a large scale photonic Recurrent Neural Network
• Wave Physics as an Analog Recurrent Neural Network
• High performance photonic reservoir computer based on a coherently driven passive cavity
• Theory of neuromorphic computing by waves: machine learning by rogue waves, dispersive shocks, and solitons
• Quantum reservoir processing
• Tutorial: Photonic Neural Networks in Delay Systems
• Filtering Random Matrices: The Effect of Incomplete Channel Control in Multiple Scattering
• Optical Reservoir Computing using multiple light scattering for chaotic systems prediction
• Random Projections through multiple optical scattering: Approximating kernels at the speed of light
• Ultrahigh enhancement of light focusing through disordered media controlled by Megapixel modes
• Large-scale spatiotemporal photonic reservoir computer for image classification
• Coupled nonlinear delay systems as deep convolutional neural networks
• Classification of time-domain waveforms using a speckle-based optical reservoir computer
• Delayed Dynamical Systems: Networks, Chimeras and Reservoir Computing
• Diffractive coupling for photonic networks: how big can we go?
• Optical transmission matrix as a probe of the photonic interaction strength
• Kernel computations from large-scale random features obtained by Optical Processing Units
• Predicting Spatio-Temporal Time Series Using Dimension Reduced Local States

Cited by in corpus (39)

• Scalable Optical Learning Operator
• Photonic extreme learning machine by free-space optical propagation
• Scalable spin-glass optical simulator
• Perspective on unconventional computing using magnetic skyrmions
• Deep Photonic Reservoir Computer Based on Frequency Multiplexing with Fully Analog Connection Between Layers
• Photonic reservoir computer based on frequency multiplexing
• Reservoir Computing Benchmarks: a tutorial review and critique
• Creating and concentrating quantum resource states in noisy environments using a quantum neural network
• Catch-22s of reservoir computing
• Machine Learning Potential of a Single Pendulum
• Characterization of multi-mode linear optical networks
• Short-wavelength Reverberant Wave Systems for Physical Realization of Reservoir Computing
• Two-color optically-addressed spatial light modulator as generic spatio-temporal systems
• Effect of temporal resolution on the reproduction of chaotic dynamics via reservoir computing
• Precursor-driven machine learning prediction of chaotic extreme pulses in Kerr resonators
• Synchronizing Chaos using Reservoir Computing
• Photonic kernel machine learning for ultrafast spectral analysis
• Beating the classical phase precision limit using a quantum neuromorphic platform
• Stochastic Reservoir Computers
• Improving the Performance of Echo State Networks Through State Feedback
• High-performance real-world optical computing trained by in situ gradient-based model-free optimization
• Quantum Next-Generation Reservoir Computing and Its Quantum Optical Implementation
• Multi-task Photonic Reservoir Computing: Wavelength Division Multiplexing for Parallel Computing with a Silicon Microring Resonator
• Optical hyperdimensional soft sensing: Speckle-based touch interface and tactile sensor
• Parallel BiLSTM-Transformer networks for forecasting chaotic dynamics
• Discovering conservation laws from trajectories via machine learning
• Photonic neural field on a silicon chip: large-scale, high-speed neuro-inspired computing and sensing
• Adiabatic evolution on a spatial-photonic Ising machine
• Linear Optical Random Projections Without Holography
• Improving the prediction of spatio-temporal chaos by combining parallel reservoir computing with dimensionality reduction
• Experimental demonstration of enhanced quantum tomography via quantum reservoir processing
• Enhanced Predictive Capability for Chaotic Dynamics by Modified Quantum Reservoir Computing
• High-Accuracy Temporal Prediction via Experimental Quantum Reservoir Computing in Correlated Spins
• Machine learning via relativity-inspired quantum dynamics
• Quantum Transport Reservoir Computing
• Phase transitions from linear to nonlinear information processing in neural networks
• Reservoir observer enhanced with residual calibration and attention mechanism
• Improvement of image classification by multiple optical scattering
• Physical reservoir computing using finitely-sampled quantum systems