High-speed analysis of spectroscopic ellipsometry data using deep learning methods

Abstract

We develop deep-learning methods for rapid analysis of spectroscopic ellipsometry data. Our approach speeds analysis by thousand-fold compared to traditional methods. We demonstrate the usefulness of our approach for a high-throughput study of phase-change alloys.