Share Email Print
cover

Proceedings Paper • new

Reinforcement learning in a large-scale photonic network (Conference Presentation)
Author(s): Daniel Brunner; Sheler Maktoobi; Louis Andreoli; Laurent Larger; Maxime Jacquot; Ingo Fischer; Julian Bueno

Paper Abstract

We experimentally create a neural network via a spatial light modulator, implementing connections between 2025 in parallel based on diffractive coupling. We numerically validate the scheme for at least 34.000 photonic neurons. Based on a digital micro-mirror array we demonstrate photonic reinforcement learning and predict a chaotic time-series via our optical neural network. The prediction error efficiently converges. Finally, we give insight based on the first investigation of effects to be encountered in neural networks physically implemented in analogue substrates.

Paper Details

Date Published: 4 March 2019
PDF
Proc. SPIE 10937, Optical Data Science II, 1093708 (4 March 2019); doi: 10.1117/12.2509351
Show Author Affiliations
Daniel Brunner, Institut Franche-Comte Electronique Mecanique Thermique et Optique (France)
CNRS (France)
Univ. Bourgogne Franche-Comte (France)
Sheler Maktoobi, Institut Franche-Comte Electronique Mecanique Thermique et Optique (France)
Louis Andreoli, Institut Franche-Comte Electronique Mecanique Thermique et Optique (France)
Laurent Larger, Institut Franche-Comte Electronique Mecanique Thermique et Optique (France)
Maxime Jacquot, Institut Franche-Comte Electronique Mecanique Thermique et Optique (France)
Ingo Fischer, Instituto de Física Interdisciplinar y Sistemas Complejos (Spain)
Consejo Superior de Investigaciones Científicas (Spain)
Univ. de les Illes Balears (Spain)
Julian Bueno, Instituto de Física Interdisciplinar y Sistemas Complejos (Spain)


Published in SPIE Proceedings Vol. 10937:
Optical Data Science II
Bahram Jalali; Ken-ichi Kitayama, Editor(s)

© SPIE. Terms of Use
Back to Top