Spectral Operator Representations

Abstract

Machine learning in atomistic materials science has grown to become a powerful tool, with most approaches focusing on atomic arrangements, typically decomposed into local atomic environments. This approach, while well-suited for machine-learned interatomic potentials, is conceptually at odds with learning complex intrinsic properties of materials, often driven by spectral properties commonly represented in reciprocal space (e.g., band gaps or mobilities) which cannot be readily atomically partitioned. For such applications, methods which represent the electronic rather than the atomic structure could be more promising. In this work, we present a general framework focused on electronic-structure descriptors which take advantage of the natural symmetries and inherent interpretability of physical models. Using this framework, we formulate two such representations and apply them respectively to measuring the similarity of carbon nanotubes and barium titanate polymorphs, and to the discovery of novel transparent conducting materials (TCMs) in the Materials Cloud 3D database (MC3D). A random forest classifier trained on 1% of the materials in the MC3D is able to correctly label 76% of entries in database which meet common screening criteria for promising TCMs.

29 pages, 8 figures, 2 tables

References in corpus (34)

• SciPy 1.0--Fundamental Algorithms for Scientific Computing in Python
• Quantum ESPRESSO: a modular and open-source software project for quantum simulations of materials
• Restoring the density-gradient expansion for exchange in solids and surfaces
• Generalized gradient approximation for solids and their surfaces
• Advanced capabilities for materials modelling with Quantum ESPRESSO
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• On representing chemical environments
• Quantum ESPRESSO toward the exascale
• Thermal Contraction and Disordering of the Al(110) Surface
• Recent developments in the PySCF program package
• Precision and efficiency in solid-state pseudopotential calculations
• DScribe: Library of Descriptors for Machine Learning in Materials Science
• Low Hole Effective Mass p-type Transparent Conducting Oxides: Identification and Design Principles
• AiiDA: Automated Interactive Infrastructure and Database for Computational Science
• The Joint Automated Repository for Various Integrated Simulations (JARVIS) for data-driven materials design
• Physics-inspired structural representations for molecules and materials
• Bandgap Change of Carbon Nanotubes: Effect of Small Tensile and Torsional Strain
• Materials Cloud, a platform for open computational science
• AiiDA 1.0, a scalable computational infrastructure for automated reproducible workflows and data provenance
• Materials Cartography: Representing and Mining Material Space Using Structural and Electronic Fingerprints
• On the Completeness of Atomic Structure Representations
• Workflows in AiiDA: Engineering a high-throughput, event-based engine for robust and modular computational workflows
• Metrics for measuring distances in configuration spaces
• A fingerprint based metric for measuring similarities of crystalline structures
• Updates to the DScribe Library: New Descriptors and Derivatives
• Representing individual electronic states for machine learning GW band structures of 2D materials
• DADApy: Distance-based Analysis of DAta-manifolds in Python
• Fermi energy determination for advanced smearing techniques
• Manifolds of quasi-constant SOAP and ACSF fingerprints and the resulting failure to machine learn four-body interactions
• Towards Novel Organic High-$T_{\mathrm{c}}$ Superconductors: Data Mining using Density of States Similarity Search
• Electronic structure based descriptor for characterizing local atomic environments
• SPA${}^{\mathrm{H}}$M: the Spectrum of Approximated Hamiltonian Matrices representations
• A unified Green's function approach for spectral and thermodynamic properties from algorithmic inversion of dynamical potentials
• Comment on "Manifolds of quasi-constant SOAP and ACSF fingerprints and the resulting failure to machine learn four body interactions"