Predicting materials properties without crystal structure: Deep representation learning from stoichiometry

Abstract

Machine learning has the potential to accelerate materials discovery by accurately predicting materials properties at a low computational cost. However, the model inputs remain a key stumbling block. Current methods typically use descriptors constructed from knowledge of either the full crystal structure -- therefore only applicable to materials with already characterised structures -- or structure-agnostic fixed-length representations hand-engineered from the stoichiometry. We develop a machine learning approach that takes only the stoichiometry as input and automatically learns appropriate and systematically improvable descriptors from data. Our key insight is to treat the stoichiometric formula as a dense weighted graph between elements. Compared to the state of the art for structure-agnostic methods, our approach achieves lower errors with less data.

arXiv commentA working implementation of our model is available at https://github.com/CompRhys/roost

References in corpus (21)

• Neural Message Passing for Quantum Chemistry
• What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?
• CALYPSO: a method for crystal structure prediction
• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• On representing chemical environments
• SchNet - a deep learning architecture for molecules and materials
• New Tolerance Factor to Predict the Stability of Perovskite Oxides and Halides
• A General-Purpose Machine Learning Framework for Predicting Properties of Inorganic Materials
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• Reconciling modern machine learning practice and the bias-variance trade-off
• Ab initio Random Structure Searching
• Big Data of Materials Science - Critical Role of the Descriptor
• Convolutional Networks on Graphs for Learning Molecular Fingerprints
• SISSO: a compressed-sensing method for identifying the best low-dimensional descriptor in an immensity of offered candidates
• Deep Ensembles: A Loss Landscape Perspective
• Alchemical and structural distribution based representation for improved QML
• Atom2Vec: learning atoms for materials discovery
• Atom-Density Representations for Machine Learning
• Bayesian semi-supervised learning for uncertainty-calibrated prediction of molecular properties and active learning
• Graph Dynamical Networks for Unsupervised Learning of Atomic Scale Dynamics in Materials
• Towards a Neural Statistician

Cited by in corpus (39)

• Recent Advances and Applications of Deep Learning Methods in Materials Science
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• Density of States Prediction for Materials Discovery via Contrastive Learning from Probabilistic Embeddings
• AtomSets -- A Hierarchical Transfer Learning Framework for Small and Large Materials Datasets
• Materials Representation and Transfer Learning for Multi-Property Prediction
• Precursor recommendation for inorganic synthesis by machine learning materials similarity from scientific literature
• Generative deep learning as a tool for inverse design of high-entropy refractory alloys
• Calibrating DFT formation enthalpy calculations by multi-fidelity machine learning
• Deep learning of experimental electrochemistry for battery cathodes across diverse compositions
• Toward Accurate Interpretable Predictions of Materials Properties within Transformer Language Models
• Extensible Structure-Informed Prediction of Formation Energy with Improved Accuracy and Usability employing Neural Networks
• A universal model for the formation energy prediction of inorganic compounds
• DenseGNN: universal and scalable deeper graph neural networks for high-performance property prediction in crystals and molecules
• Artificial Intelligence in Materials Science and Engineering: Current Landscape, Key Challenges, and Future Trajectorie
• Robust model benchmarking and bias-imbalance in data-driven materials science: a case study on MODNet
• High-dimensional Bayesian Optimization of Hyperparameters for an Attention-based Network to Predict Materials Property: a Case Study on CrabNet using Ax and SAASBO
• Formula graph self-attention network for representation-domain independent materials discovery
• Random projections and Kernelised Leave One Cluster Out Cross-Validation: Universal baselines and evaluation tools for supervised machine learning for materials properties
• Accelerated chemical space search using a quantum-inspired cluster expansion approach
• Ab initio machine learning of phase space averages
• Machine Learning for compositional disorder: A Comparison Between Different Descriptors and Machine Learning Frameworks
• Artificial Intelligence for Science in Quantum, Atomistic, and Continuum Systems
• Materials Graph Transformer predicts the outcomes of inorganic reactions with reliable uncertainties
• Closed-loop Error Correction Learning Accelerates Experimental Discovery of Thermoelectric Materials
• Coarse-grained crystal graph neural networks for reticular materials design
• Crystal structure prediction with host-guided inpainting generation and foundation potentials
• Si plate radius influence on the photoacoustic signal processed by neural networks
• Adaptive Catalyst Discovery Using Multicriteria Bayesian Optimization with Representation Learning
• Active learning based generative design for the discovery of wide bandgap materials
• Neural Structure Fields with Application to Crystal Structure Autoencoders
• Machine learning for accelerated bandgap prediction in strain-engineered quaternary III-V semiconductors
• XtalOpt Version 14: Variable-Composition Crystal Structure Search for Functional Materials Through Pareto Optimization
• Beyond Predicted ZT: Machine Learning Strategies for the Experimental Discovery of Thermoelectric Materials
• CHILI: Chemically-Informed Large-scale Inorganic Nanomaterials Dataset for Advancing Graph Machine Learning
• MLatticeABC: Generic Lattice Constant Prediction of Crystal Materials using Machine Learning
• Elemental Reactivity Maps for Materials Discovery
• Computational discovery of new 2D materials using deep learning generative models
• Electron-Informed Coarse-Graining Molecular Representation Learning for Real-World Molecular Physics
• Formation Energy Prediction of Material Crystal Structures using Deep Learning