Reservoir computing with a single time-delay autonomous Boolean node

Haynes, Nicholas D.; Soriano, Miguel C.; Rosin, David P.; Fischer, Ingo; Gauthier, Daniel J.

doi:10.1103/physreve.91.020801

Cited by 116 publications

(74 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such systems can be exploited to realise a form of analog computer based on the Reservoir Computing (RC) paradigm [9,10] in which unoptimised high-dimensional dynamic systems (termed reservoirs) are used as signal processors. The RC approach is simple, versatile and can be applied to a wide set of problems (see the review [11]) and experimental implementations [12][13][14][15][16][17][18][19][20]. Applying the BP algorithm to delay-coupled signal processors allows one to optimise many more parameters than in traditional RC, yielding significant improvements in performance as was shown in simulation in [4], and subsequently in an experiment [5] in which BP was applied to a numerical model of the system, and the results of the BP algorithm applied to the physical experimental setup.…”

mentioning

confidence: 99%

Embodiment of Learning in Electro-Optical Signal Processors

et al. 2016

View full text Add to dashboard Cite

Delay-coupled electro-optical systems have received much attention for their dynamical properties and their potential use in signal processing. In particular it has recently been demonstrated, using the artificial intelligence algorithm known as reservoir computing, that photonic implementations of such systems solve complex tasks such as speech recognition. Here we show how the backpropagation algorithm can be physically implemented on the same electro-optical delay-coupled architecture used for computation with only minor changes to the original design. We find that, compared when the backpropagation algorithm is not used, the error rate of the resulting computing device, evaluated on three benchmark tasks, decreases considerably. This demonstrates that electro-optical analog computers can embody a large part of their own training process, allowing them to be applied to new, more difficult tasks.Published version : http://dx

show abstract

mentioning

confidence: 99%

Embodiment of Learning in Electro-Optical Signal Processors

et al. 2016

View full text Add to dashboard Cite

show abstract

“…Because of its structural simplicity, reservoir computers have motivated many hardware implementations ranging from electronics [14], [15] and spintronics [16], [17], to photonics [18], [19] with the goal of reaching ultra-fast processing speed with an energy consumption at least two order of magnitudes below that of a software-based RC running on a computer. Unprecedented classification speed have been recently achieved on the spoken-digit recognition task from the TIMIT46 data base with real-time processing speed ranging from 300,000 to 1,000,000 words analyzed per second using laser with optical feedback [19] and optoelectronic oscillators arXiv:2004.02542v1 [cs.NE] 6 Apr 2020 [20], respectively.…”

Section: Introductionmentioning

confidence: 99%

Large-Scale Spatiotemporal Photonic Reservoir Computer for Image Classification

Antonik

Marsal²,

Rontani³

2020

IEEE J. Select. Topics Quantum Electron.

View full text Add to dashboard Cite

We propose a scalable photonic architecture for implementation of feedforward and recurrent neural networks to perform the classification of handwritten digits from the MNIST database. Our experiment exploits off-the-shelf optical and electronic components to currently achieve a network size of 16,384 nodes. Both network types are designed within the the reservoir computing paradigm with randomly weighted input and hidden layers. Using various feature extraction techniques (e.g. histograms of oriented gradients, zoning, Gabor filters) and a simple training procedure consisting of linear regression and winner-takes-all decision strategy, we demonstrate numerically and experimentally that a feedforward network allows for classification error rate of 1%, which is at the state-of-the-art for experimental implementations and remains competitive with more advanced algorithmic approaches. We also investigate recurrent networks in numerical simulations by explicitly activating the temporal dynamics, and predict a performance improvement over the feedforward configuration.

show abstract

“…Field Programmable Gate Arrays (FPGAs) are particularly suited for such applications. Several RC implementations using an FPGA chip have been reported [2,8,17], but no interconnection between an FPGA and a physical reservoir computer has been reported so far.…”

Section: Introductionmentioning

confidence: 99%

Online Training of an Opto-Electronic Reservoir Computer

Antonik

Duport

Smerieri

et al. 2015

Neural Information Processing

View full text Add to dashboard Cite

Abstract. Reservoir Computing is a bio-inspired computing paradigm for processing time dependent signals. Its analog implementations equal and sometimes outperform other digital algorithms on a series of benchmark tasks. Their performance can be increased by switching from offline to online training method. Here we present the first online trained opto-electronic reservoir computer. The system is tested on a channel equalisation task and the algorithm is executed by an FPGA chip. We report performances close to previous implementations and demonstrate the benefits of online training on a non-stationary task that could not be easily solved using offline methods.

show abstract

Reservoir computing with a single time-delay autonomous Boolean node

Cited by 116 publications

References 22 publications

Embodiment of Learning in Electro-Optical Signal Processors

Embodiment of Learning in Electro-Optical Signal Processors

Large-Scale Spatiotemporal Photonic Reservoir Computer for Image Classification

Online Training of an Opto-Electronic Reservoir Computer

Contact Info

Product

Resources

About