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.
Highlights
Recent studies in machine learning show that large neural networks can dramatically improve the network performance; their realization with conventional computing technologies is to date a significant challenge
reservoir computing (RC) is a relatively recent computational framework [1,2] derived from independently proposed recurrent neural network (RNN) models, such as echo state networks (ESNs) [3] and liquid state machines (LSMs) [4]
Before starting the actual experiment, we perform a grid search to optimize a set of tunable parameters in our optical scheme
Summary
Recent studies in machine learning show that large neural networks can dramatically improve the network performance; their realization with conventional computing technologies is to date a significant challenge. The first approach is based on a single nonlinear node with a time-delayed optoelectronic or all-optical feedback in order to get time-multiplexed virtual nodes in the temporal domain [12,13,14,15,16,17,18,19,20,21,22,23,24] Such architectures can reach supercomputer performances, e.g., gigabyte per second data rates for chaotic time-series prediction tasks [25] or million words per second classification for speech recognition tasks [26]. Another popular approach of photonic RC is based on spatially distributed nonlinear nodes The latter is endowed by its intrinsic property to process large-scale input information without sacrificing the computation speed. Our results are hardly reachable by the conventional von Neumann machines, and they significantly advance the state of the art of the unconventional reservoir computing approaches, in general
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
