Symbolic regression in nano-optics: characterization of dispersive materials as a case study

Macias, Demetrio; Canales, Claudio


In this contribution we propose a Symbolic Regression (SR) scheme, based on Genetic Programming (GP), to retrieve the closed expression that represents the dispersion model of a given material. To this end, we consider an isotropic, homogeneous and lossless dispersive dielectric and assess, through some examples, the possibilities and limitations of our approach when used to solve this kind of inverse prob- lem. Furthermore, we discuss its potential applications in other research fields as, for example, nano-photonics or plasmonics.

Más información

Fecha de publicación: 2021