Photonic Reservoir Computing: Optical Recurrent Neural Networks

by Daniel Brunner (Editor), Miguel C. Soriano (Editor), Guy Van der Sande (Editor)

Overview

Photonics has long been considered an attractive substrate for next generation implementations of machine-learning concepts. Reservoir Computing tremendously facilitated the realization of recurrent neural networks in analogue hardware. This concept exploits the properties of complex nonlinear dynamical systems, giving rise to photonic reservoirs implemented by semiconductor lasers, telecommunication modulators and integrated photonic chips.

Product Details

ISBN-13:	9783110582000
Publisher:	De Gruyter
Publication date:	07/08/2019
Pages:	277
Product dimensions:	6.69(w) x 9.45(h) x (d)
Age Range:	18 Years

About the Author

Daniel Brunner, U Besancon, France;
Miguel C. Sariano, U Illes Balears, Spain;
Guy Van der Sande, U Brussels, Belgium.

Table of Contents

Preface v

List of Contributing Authors xi

1 Introduction to novel photonic computing Daniel Brunner Piotr Antonik Xavier Porte 1

2 Information processing and computation with photonic reservoir systems Joni Dambre 33

3 Integrated on-chip reservoirs Andrew Katumba Matthias Freiberger Floris Laporte Alessio Lugnan Stijn Sackesyn Chonghuai Ma Joni Dambre Peter Bienstman 53

4 Large scale spatiotemporal reservoirs Daniel Brunner Julian Bueno Xavier Porte Sheler Maktoobi Louis Andreoli 83

5 Time delay systems for reservoir computing Silvia Ortín Luis Pesquera Guy Van der Sande Miguel C. Soriano 117

6 Ikeda delay dynamics as Reservoir processors Laurent Larger 153

7 Semiconductor lasers as reservoir substrates Guy Van der Sande Miguel C. Soriano 185

8 Advanced reservoir computers: analogue autonomous systems and real time control Piotr Antonik Serge Massar François Duport 205

Outlook 259

Index 263

Editorial Reviews

"The book is very clearly written by experts in the field and contains numerous experiments and references. It can be used by anyone working in optical computing, which has developed very much in recent years due to optical implementations of neural networks."
Mircea Dragoman in: OSA. The Optic Society (02.04.2020), https://www.osa-opn.org/home/book_reviews/2020/0420/photonic_reservoir_computing_optical_recurrent_neu/

From the Publisher