MODNet -- accurate and interpretable property predictions for limited materials datasets by feature selection and joint-learning

Abstract

In order to make accurate predictions of material properties, current machine-learning approaches generally require large amounts of data, which are often not available in practice. In this work, an all-round framework is presented which relies on a feedforward neural network, the selection of physically-meaningful features and, when applicable, joint-learning. Next to being faster in terms of training time, this approach is shown to outperform current graph-network models on small datasets. In particular, the vibrational entropy at 305 K of crystals is predicted with a mean absolute test error of 0.009 meV/K/atom (four times lower than previous studies). Furthermore, joint-learning reduces the test error compared to single-target learning and enables the prediction of multiple properties at once, such as temperature functions. Finally, the selection algorithm highlights the most important features and thus helps understanding the underlying physics.

5 pages, 2 figures

References in corpus (3)

• Big Data of Materials Science - Critical Role of the Descriptor
• AtomSets -- A Hierarchical Transfer Learning Framework for Small and Large Materials Datasets
• Automation methodologies and large-scale validation for $GW$, towards high-throughput $GW$ calculations

Cited by in corpus (14)

• A critical examination of robustness and generalizability of machine learning prediction of materials properties
• AtomSets -- A Hierarchical Transfer Learning Framework for Small and Large Materials Datasets
• Developments and applications of the OPTIMADE API for materials discovery, design, and data exchange
• An equivariant graph neural network for the elasticity tensors of all seven crystal systems
• Hierarchical symbolic regression for identifying key physical parameters correlated with bulk properties of perovskites
• A Quantum-Chemical Bonding Database for Solid-State Materials
• 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
• Classification-based detection and quantification of cross-domain data bias in materials discovery
• Optical materials discovery and design with federated databases and machine learning
• Accelerating the discovery of high-performance nonlinear optical materials using active learning and high-throughput screening
• Superconductor discovery in the emerging paradigm of Materials Informatics
• Benchmarking Inverse Optimization Algorithms for Materials Design