Chen Chi, Zuo Yunxing, Ye Weike, Li Xiangguo, Ong Shyue Ping
Department of NanoEngineering, University of California, San Diego, CA, USA.
Nat Comput Sci. 2021 Jan;1(1):46-53. doi: 10.1038/s43588-020-00002-x. Epub 2021 Jan 14.
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.
从材料原子排列预测其性质是材料科学的一个基本目标。近年来，机器学习作为一种提供材料性质快速预测的新范式出现了，但其实际效用受到高保真数据稀缺的限制。在这里，我们开发了多保真度图网络，作为一种通用方法，以在小数据量的情况下实现对材料性质的准确预测。作为概念验证，我们表明，纳入低保真的佩德韦-伯克-恩泽霍夫带隙极大地提高了材料图中潜在结构特征的分辨率，导致实验带隙预测的平均绝对误差降低了22%-45%。我们进一步证明，材料图网络中学习到的元素嵌入为材料中的无序建模提供了一种自然方法，解决了材料性质计算预测中的一个基本差距。
