Benchmarking graph neural networks for materials chemistry

By Victor Fung, Jiaxin Zhang, Eric J. Juarez, Bobby G. Sumpter. npj Computational Materials, 2021

313 citations recorded by the source corpus.

Graph neural networks (GNNs) have received intense interest as a rapidly expanding class of machine learning models remarkably well-suited for materials applications. To date, a number of successful GNNs have been proposed and demonstrated for systems ranging from crystal stability to electronic property prediction and to surface chemistry and heterogeneous catalysis. However, a consistent benchmark of these models remains lacking, hindering the development and consistent evaluation of new models in the materials field. Here, we present a workflow and testing platform, MatDeepLearn, for quickly and reproducibly assessing and comparing GNNs and other machine learning models. We use this platform to optimize and evaluate a selection of top performing GNNs on several representative datasets in computational materials chemistry. From our investigations we note the importance of hyperparameter selection and find roughly similar performances for the top models once optimized. We identify several strengths in GNNs over conventional models in cases with compositionally diverse datasets and in its overall flexibility with respect to inputs, due to learned rather than defined representations. Meanwhile several weaknesses of GNNs are also observed including high data requirements, and suggestions for further improvement for applications in materials chemistry are discussed.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Materials Chemistry

View Paper: https://www.semanticscholar.org/paper/d08167fd8583b0f70ba8a26821c29ea8af420826

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xianglin Liu
2. Zongrui Pei
3. Jiaxin Zhang

In Computational Theory and Mathematics

1. Xianglin Liu
2. Zongrui Pei
3. Jiaxin Zhang

In Materials Chemistry

1. S. Dai
2. Yifan Sun
3. Xianglin Liu
4. Zongrui Pei
5. Jiaxin Zhang

Citations in the corpus

• Machine Learning for High-entropy Alloys: Progress, Challenges and Opportunities (2022, methodology): “These results were summarized in Ref.”
• High-entropy materials for catalysis: A new frontier (2021, background): “In addition, different algorithms should be adopted for different purposes (143).”
• ChemTEB: Chemical Text Embedding Benchmark, an Overview of Embedding Models Performance & Efficiency on a Specific Domain (2024, unknown): “MatDeepLearn [54] benchmarks GNNs for materials chemistry applications, identifying strengths in handling compositionally diverse datasets but also noting the high data requirements and limitations.”