Transfer learning for photonic delay-based reservoir computing to compensate parameter drift.
| Autor: | Bauwens I; Applied Physics Research Group, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium., Harkhoe K; Applied Physics Research Group, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium., Bienstman P; Photonics Research Group, Department of Information Technology, Ghent University-IMEC, Technologiepark Zwijnaarde 126, 9052 Ghent, Belgium., Verschaffelt G; Applied Physics Research Group, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium., Van der Sande G; Applied Physics Research Group, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium. |
|---|---|
| Jazyk: | angličtina |
| Zdroj: | Nanophotonics (Berlin, Germany) [Nanophotonics] 2022 Oct 18; Vol. 12 (5), pp. 949-961. Date of Electronic Publication: 2022 Oct 18 (Print Publication: 2023). |
| DOI: | 10.1515/nanoph-2022-0399 |
| Abstrakt: | Photonic reservoir computing has been demonstrated to be able to solve various complex problems. Although training a reservoir computing system is much simpler compared to other neural network approaches, it still requires considerable amounts of resources which becomes an issue when retraining is required. Transfer learning is a technique that allows us to re-use information between tasks, thereby reducing the cost of retraining. We propose transfer learning as a viable technique to compensate for the unavoidable parameter drift in experimental setups. Solving this parameter drift usually requires retraining the system, which is very time and energy consuming. Based on numerical studies on a delay-based reservoir computing system with semiconductor lasers, we investigate the use of transfer learning to mitigate these parameter fluctuations. Additionally, we demonstrate that transfer learning applied to two slightly different tasks allows us to reduce the amount of input samples required for training of the second task, thus reducing the amount of retraining. (© 2022 the author(s), published by De Gruyter, Berlin/Boston.) |
| Databáze: | MEDLINE |
| Externí odkaz: |
|