Learning Properties of Ordered and Disordered Materials from Multi-fidelity Data

Abstract

Predicting the properties of a material from the arrangement of its atoms is a fundamental goal in materials science. While machine learning has emerged in recent years as a new paradigm to provide rapid predictions of materials properties, their practical utility is limited by the scarcity of high-fidelity data. Here, we develop multi-fidelity graph networks as a universal approach to achieve accurate predictions of materials properties with small data sizes. As a proof of concept, we show that the inclusion of low-fidelity Perdew-Burke-Ernzerhof band gaps greatly enhances the resolution of latent structural features in materials graphs, leading to a 22-45\% decrease in the mean absolute errors of experimental band gap predictions. We further demonstrate that learned elemental embeddings in materials graph networks provide a natural approach to model disorder in materials, addressing a fundamental gap in the computational prediction of materials properties.

References in corpus (9)

• Gaussian Approximation Potentials: the accuracy of quantum mechanics, without the electrons
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• SchNet - a deep learning architecture for molecules and materials
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• A Performance and Cost Assessment of Machine Learning Interatomic Potentials
• Big Data meets Quantum Chemistry Approximations: The $\Delta$-Machine Learning Approach
• A composite neural network that learns from multi-fidelity data: Application to function approximation and inverse PDE problems
• Kohn-Sham potential with discontinuity for band gap materials
• A Community Contribution Framework for Sharing Materials Data with Materials Project

Cited by in corpus (21)

• Recent Advances and Applications of Deep Learning Methods in Materials Science
• A Universal Graph Deep Learning Interatomic Potential for the Periodic Table
• Multi-fidelity Bayesian Neural Networks: Algorithms and Applications
• AtomSets -- A Hierarchical Transfer Learning Framework for Small and Large Materials Datasets
• Materials Representation and Transfer Learning for Multi-Property Prediction
• Calibrating DFT formation enthalpy calculations by multi-fidelity machine learning
• A rule-free workflow for the automated generation of databases from scientific literature
• Deciphering Chemical Ordering in High Entropy Materials: A Machine Learning-Accelerated High-throughput Cluster Expansion Approach
• Robust model benchmarking and bias-imbalance in data-driven materials science: a case study on MODNet
• Constructing and evaluating machine-learned interatomic potentials for Li-based disordered rocksalts
• Decoding Structure-Spectrum Relationships with Physically Organized Latent Spaces
• Reconstruction of nanoscale particles from single-shot wide-angle FEL diffractions patterns with physics-informed neural networks
• Self-supervised graph neural networks for accurate prediction of Néel temperature
• Space Group Informed Transformer for Crystalline Materials Generation
• Learning the Crystal Structure Genome for Property Classification
• Multi-Objective Quality-Diversity for Crystal Structure Prediction
• Emerging Microelectronic Materials by Design: Navigating Combinatorial Design Space with Scarce and Dispersed Data
• Investigating Data Hierarchies in Multifidelity Machine Learning for Excitation Energies
• Data Fusion of Deep Learned Molecular Embeddings for Property Prediction
• Reinforcement Fine-Tuning for Materials Design
• Scalable Data-Driven Basis Selection for Linear Machine Learning Interatomic Potentials