Machine Learning Reconstruction of High-dimensional Electronic Structure from Angle-resolved Photoemission Spectroscopy

Extracting key electronic parameters from complex angle-resolved photoemission spectroscopy data is a significant challenge in condensed matter physics. This research introduces an advanced machine learning method, utilizing implicit neural representations, to automate and accelerate this process. Our model is trained to learn the direct relationship between a material's fundamental electronic parameters and its high-dimensional ARPES spectra. Applied to perovskite nickelates, this approach successfully obtained a more precise set of parameters from experimental data, outperforming traditional analytical techniques. This work demonstrates the power of implicit neural representations to bridge the gap between theory and experiment, paving the way for high-throughput discovery in materials science.