AtomSets -- A Hierarchical Transfer Learning Framework for Small and Large Materials Datasets

Abstract

Predicting materials properties from composition or structure is of great interest to the materials science community. Deep learning has recently garnered considerable interest in materials predictive tasks with low model errors when dealing with large materials data. However, deep learning models suffer in the small data regime that is common in materials science. Here we leverage the transfer learning concept and the graph network deep learning framework and develop the AtomSets machine learning framework for consistent high model accuracy at both small and large materials data. The AtomSets models can work with both compositional and structural materials data. By combining with transfer learned features from graph networks, they can achieve state-of-the-art accuracy from using small compositional data (<400) to large structural data (>130,000). The AtomSets models show much lower errors than the state-of-the-art graph network models at small data limits and the classical machine learning models at large data limits. They also transfer better in the simulated materials discovery process where the targeted materials have property values out of the training data limits. The models require minimal domain knowledge inputs and are free from feature engineering. The presented AtomSets model framework opens new routes for machine learning-assisted materials design and discovery.

References in corpus (11)

• Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties
• Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals
• Predicting materials properties without crystal structure: Deep representation learning from stoichiometry
• Kohn-Sham potential with discontinuity for band gap materials
• Developing an improved Crystal Graph Convolutional Neural Network framework for accelerated materials discovery
• A critical examination of compound stability predictions from machine-learned formation energies
• Learning Properties of Ordered and Disordered Materials from Multi-fidelity Data
• MODNet -- accurate and interpretable property predictions for limited materials datasets by feature selection and joint-learning
• Overcoming data scarcity with transfer learning
• IRNet: A General Purpose Deep Residual Regression Framework for Materials Discovery
• Accelerating Materials Discovery with Bayesian Optimization and Graph Deep Learning

Cited by in corpus (7)

• A critical examination of robustness and generalizability of machine learning prediction of materials properties
• MODNet -- accurate and interpretable property predictions for limited materials datasets by feature selection and joint-learning
• Advances of Machine Learning in Materials Science: Ideas and Techniques
• Formula graph self-attention network for representation-domain independent materials discovery
• Hardness and fracture toughness models by symbolic regression
• Optical materials discovery and design with federated databases and machine learning
• Self-supervised Representations and Node Embedding Graph Neural Networks for Accurate and Multi-scale Analysis of Materials