Bayesian optimization with improved scalability and derivative information for efficient design of nanophotonic structures
| Autor: | Carsten Rockstuhl, Philipp-Immanuel Schneider, Sven Burger, Xavier Garcia-Santiago |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Optimal design
FOS: Computer and information sciences Computer science Bayesian optimization Bayesian probability FOS: Physical sciences Machine Learning (stat.ML) 02 engineering and technology Parameter space Computational Physics (physics.comp-ph) Atomic and Molecular Physics and Optics Range (mathematics) 020210 optoelectronics & photonics Statistics - Machine Learning Scalability 0202 electrical engineering electronic engineering information engineering Point (geometry) Enhanced Data Rates for GSM Evolution Algorithm Physics - Computational Physics Physics - Optics Optics (physics.optics) |
| DOI: | 10.48550/arxiv.2101.02972 |
| Popis: | We propose the combination of forward shape derivatives and the use of an iterative inversion scheme for Bayesian optimization to find optimal designs of nanophotonic devices. This approach widens the range of applicability of Bayesian optmization to situations where a larger number of iterations is required and where derivative information is available. This was previously impractical because the computational efforts required to identify the next evaluation point in the parameter space became much larger than the actual evaluation of the objective function. We demonstrate an implementation of the method by optimizing a waveguide edge coupler. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|