Photonic single nonlinear-delay dynamical node for information processing

Silvia Ortín; Daniel San-Martín; Luis Pesquera; José Manuel Gutiérrez

doi:10.1117/12.922463

10 May 2012 Photonic single nonlinear-delay dynamical node for information processing

Silvia Ortín, Daniel San-Martín, Luis Pesquera, José Manuel Gutiérrez

Author Affiliations +

Proceedings Volume 8432, Semiconductor Lasers and Laser Dynamics V; 843214 (2012) https://doi.org/10.1117/12.922463
Event: SPIE Photonics Europe, 2012, Brussels, Belgium

Abstract

An electro-optical system with a delay loop based on semiconductor lasers is investigated for information processing by performing numerical simulations. This system can replace a complex network of many nonlinear elements for the implementation of Reservoir Computing. We show that a single nonlinear-delay dynamical system has the basic properties to perform as reservoir: short-term memory and separation property. The computing performance of this system is evaluated for two prediction tasks: Lorenz chaotic time series and nonlinear auto-regressive moving average (NARMA) model. We sweep the parameters of the system to find the best performance. The results achieved for the Lorenz and the NARMA-10 tasks are comparable to those obtained by other machine learning methods.

Citation Download Citation

Silvia Ortín, Daniel San-Martín, Luis Pesquera, and José Manuel Gutiérrez "Photonic single nonlinear-delay dynamical node for information processing", Proc. SPIE 8432, Semiconductor Lasers and Laser Dynamics V, 843214 (10 May 2012); https://doi.org/10.1117/12.922463

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE