Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This study explores resonance prediction in dispersive media by formulating the problem as nonlinear spectral issues involving volume integral operators. By leveraging asymptotic analysis, the authors not only establish a baseline approximation but also guide the development of predictive correction models that learn the residuals between asymptotic and reference resonances using features derived from subwavelength expansions. The approach yields significant improvements in predictions for single resonators and dimers, with symbolic regression generating compact, interpretable formulas for the learned residuals.
Asymptotic analysis transforms resonance prediction from a mere approximation tool into a powerful guide for developing accurate and interpretable correction models.
We study resonance prediction in dispersive media, formulated as nonlinear spectral problems for volume integral operators. The main idea is to use asymptotic analysis not only as a baseline approximation, but also as a guide for constructing predictive correction models. We learn the residual between asymptotic and reference resonances using features suggested by the subwavelength expansion, including the logarithmic scales specific to two dimensions. The resulting corrections substantially improve single-resonator and dimer predictions, and symbolic regression produces compact formulas for the learned residual. The results show that asymptotic analysis can be used not only to approximate resonances, but also to design the feature space in which data-driven corrections become accurate, low-dimensional, and interpretable.