Beyond the conventional trial-and-error method, machine learning offers a great opportunity to accelerate the discovery of functional materials, but still often suffers from difficulties such as limited materials data and unbalanced distribution of target property. Here, we propose the ab initio Bayesian active learning method that combines active learning and high-throughput ab initio calculations to accelerate prediction of desired functional materials with the ultrahigh efficiency and accuracy. We apply it as an instance to a large family (3,119) of two-dimensional hexagonal binary compounds with unbalanced materials property, and accurately screen out the materials with maximal electric polarization and proper photovoltaic band gaps, respectively, whereas the computational costs are significantly reduced by only calculating a few tenths of possible candidates in comparison to the random search. This approach shows enormous advantages for the cases with unbalanced distributions of target property. It can be readily applied to seek for a broad range of advanced materials.
Active Learning for Predicting the Enthalpy of Mixing inBinary Liquids Based on Ab Initio Molecular Dynamics
Univ. Grenoble Alpes, CNRS, Grenoble INP, SIMaP(格勒诺布尔阿尔卑斯大学,法国国家科学研究中心,格勒诺布尔理工学院,材料科学与物理实验室);Interdisciplinary Centre for Advanced Materials Simulations (ICAMS), Ruhr Universität Bochum(鲁尔大学波鸿分校先进材料模拟跨学科中心);Center for Basic Research on Materials, National Institute for Materials Science(物质材料研究机构基础材料研究中心);Graduate School of Frontier Sciences, The University of Tokyo(东京大学前沿科学研究生院)
Active-Learning Inspired $\textit{Ab Initio}$ Theory-Experiment Loop Approach for Management of Material Defects: Application to Superconducting Qubits
Cornell University(康奈尔大学);Syracuse University(雪城大学);Massachusetts Institute of Technology(麻省理工学院)
Towards Extended Active Learning for Modelling Ferroelectric Switching: the Need for 'Gold Standards'
International Centre for Quantum and Molecular Structures and the School of Physics, Shanghai University(上海大学量子与分子结构国际中心及物理学院);University of Technology Sydney(悉尼科技大学);The University of Queensland(昆士兰大学);Griffith University(格里菲斯大学);Materials Genome Institute, International Centre for Quantum and Molecular Structures, Shanghai University(上海大学量子与分子结构国际中心材料基因组研究院)
Machine Learning Inter-Atomic Potentials Generation Driven by Active Learning: A Case Study for Amorphous and Liquid Hafnium dioxide
Leadership Computing Facility, Argonne National Laboratory(阿贡国家实验室领导计算设施);Institute for Computational Physics, Universität Stuttgart(斯图加特大学计算物理研究所);Applied Materials Division, Argonne National Laboratory(阿贡国家实验室应用材料部);University of Cambridge(剑桥大学);X-ray Science Division, Argonne National Laboratory(阿贡国家实验室X射线科学部);Computational Science Division, Argonne National Laboratory(阿贡国家实验室计算科学部)
Accelerating the Training and Improving the Reliability of Machine-Learned Interatomic Potentials for Strongly Anharmonic Materials through Active Learning