Published August 31, 2026 | Version v1
Software Open

ERI-Net: Machine Learning Inverse Design of DBR Cavities with Complex Index Materials

Authors/Creators

  • 1. ROR icon University of Dayton

Description

ERINet is a neural network trained to predict active and phase compensated layer thicknesses (tC, tD) in a multilayer DBR cavity , given design parameters (N, M, n', κ, λ, nH, nL). This upload includes the training dataset, trained model, code for training, validation, parametric sweep analysis, and figure generation.

Files

Files (3.1 MB)

Name Size
md5:d2935fcc4223e76afcd6af505bffef05
23.6 kB Download
md5:5c75db82092412ec5d218c80e510fd83
2.2 MB Download
md5:0617702d470df9dbf9ec70aa7500d6d4
143.6 kB Download
md5:b8b648c80be920103a04d67fc006219b
640.1 kB Download
md5:4e4b93a7322cf91ce2b12b186cc9927b
53.2 kB Download

Additional details

Dates

Created
2026-08-30