Indexed by: Journal paper

Journal: Physical Review Applied

Affiliation of Author(s): 光电学院，国家光电研究中心

Place of Publication: 美国

Discipline: 物理学

Funded by: 自然科学基金

Document Type: J

Volume: 14

Issue: 4

Page Number: 044032

Key Words: 无

DOI number: 10.1103/PhysRevApplied.14.044032

Date of Publication: 2020-09-08

Teaching and Research Group: c716

Abstract: Neural networks based on machine learning can interpolate well within the training dataset, but their ability to extrapolate is severely limited by fundamental issues such as the bias-variance trade-off. Here we introduce the concept of an operator parameter space consisting of physical entities encoded with Maxwell’s equations to improve the networks’ capability to generalize beyond their training set. We illustrate the idea with photonic crystals, and show that the network trained with operator parameters yields remarkably accurate predictions of the topological transitions both within and beyond the training physical space. Such concepts can be generalized to higher-dimensional wave structures by choosing the appropriate operator parameters.

Note: 无

Links to published journals: https://doi.org/10.1103/PhysRevApplied.14.044032