SOA-based reservoir computing

A. Bednyakova*, E. Manuylovich, D. A. Ivoilov, I. S. Terekhov, S. K. Turitsyn

*Corresponding author for this work

Research output: Chapter in Book/Published conference outputConference publication

Abstract

We present a novel reservoir computing approach in which single nonlinear device - semiconductor optical amplifier, replaces the entire nonlinear reservoir to perform computations, without the conventional use of a delay loop. To study the performance of the proposed scheme, we use it for the benchmark prediction task of learning the Mackey-Glass chaotic attractor.

Original languageEnglish
Title of host publication2024 International Conference Laser Optics (ICLO)
PublisherIEEE
Pages279
Number of pages1
ISBN (Electronic)9798350390674
DOIs
Publication statusE-pub ahead of print - 19 Aug 2024
Event2024 International Conference Laser Optics, ICLO 2024 - St. Petersburg, Russian Federation
Duration: 1 Jul 20245 Jul 2024

Publication series

NameProceedings of International Conference Laser Optics (ICLO)

Conference

Conference2024 International Conference Laser Optics, ICLO 2024
Country/TerritoryRussian Federation
CitySt. Petersburg
Period1/07/245/07/24

Keywords

  • linear regression
  • Mackey-Glass time series
  • optical up-sampling
  • reservoir computing
  • semiconductor optical amplifier

Fingerprint

Dive into the research topics of 'SOA-based reservoir computing'. Together they form a unique fingerprint.

Cite this