Representing individual electronic states for machine learning GW band structures of 2D materials

Abstract

Choosing optimal representation methods of atomic and electronic structures is essential when machine learning properties of materials. We address the problem of representing quantum states of electrons in a solid for the purpose of machine leaning state-specific electronic properties. Specifically, we construct a fingerprint based on energy decomposed operator matrix elements (ENDOME) and radially decomposed projected density of states (RAD-PDOS), which are both obtainable from a standard density functional theory (DFT) calculation. Using such fingerprints we train a gradient boosting model on a set of 46k G${}_0$W${}_0$ quasiparticle energies. The resulting model predicts the self-energy correction of states in materials not seen by the model with a mean absolute error of 0.14 eV. By including the material's calculated dielectric constant in the fingerprint the error can be further reduced by 30%, which we find is due to an enhanced ability to learn the correlation/screening part of the self-energy. Our work paves the way for accurate estimates of quasiparticle band structures at the cost of a standard DFT calculation.

References in corpus (8)

• A General-Purpose Machine Learning Framework for Predicting Properties of Inorganic Materials
• Big Data of Materials Science - Critical Role of the Descriptor
• Renormalization of Molecular Electronic Levels at Metal-Molecule Interfaces
• Recent Progress of the Computational 2D Materials Database (C2DB)
• Probing excitonic states in ultraclean suspended two-dimensional semiconductors by photocurrent spectroscopy
• The GW compendium: A practical guide to theoretical photoemission spectroscopy
• Simple vertex correction improves GW band energies of bulk and two-dimensional crystals
• Crystal Structure Representations for Machine Learning Models of Formation Energies

Cited by in corpus (9)

• Multimodal Foundation Models for Material Property Prediction and Discovery
• Machine learning approach to genome of two-dimensional materials with flat electronic bands
• Unified Deep Learning Framework for Many-Body Quantum Chemistry via Green's Functions
• Similarity of materials and data-quality assessment by fingerprinting
• Ab initio property characterisation of thousands of previously unknown 2D materials
• Elf autoencoder: unsupervised exploration of flat-band materials using electronic band structure fingerprints
• Spectral Operator Representations
• MADAS -- A Python framework for assessing similarity in materials-science data
• Electron correlation in semiconductors and insulators via symbolic regression