Generalization Properties of Machine Learning-based Raman Models

U. C. De Moura, D. Zibar, A. M. Rosa Brusin, A. Carena, F. Da Ros

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

We investigate the generalization capabilities of neural network-based Raman amplifier models. The new proposed model architecture, including fiber parameters as inputs, can predict Raman gains of fiber types unseen during training, unlike previous fiber-specific models.

Original languageEnglish
Title of host publicationProceedings of 2021 Optical Fiber Communications Conference and Exhibition
PublisherIEEE
Publication dateJun 2021
Article number9489950
ISBN (Electronic)9781943580866
Publication statusPublished - Jun 2021
Event2021 Optical Fiber Communications Conference and Exhibition - San Francisco, United States
Duration: 6 Jun 202111 Jun 2021

Conference

Conference2021 Optical Fiber Communications Conference and Exhibition
Country/TerritoryUnited States
CitySan Francisco
Period06/06/202111/06/2021

Bibliographical note

Funding Information:
We thank OFS Fitel Denmark for providing the fibers used in this work. This project has received funding from the European Research Council through the ERC-CoG FRECOM project (grant agreement no. 771878), the European Union’s Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 754462 and the Villum Foundations (VYI OPTIC-AI grant no. 29344).

Funding Information:
We thank OFS Fitel Denmark for providing the fibers used in this work.

Publisher Copyright:
© 2021 OSA.

Fingerprint

Dive into the research topics of 'Generalization Properties of Machine Learning-based Raman Models'. Together they form a unique fingerprint.

Cite this