Materials property prediction for limited datasets enabled by feature selection and joint learning with MODNet

作     者：Pierre-Paul De Breuck Geoffroy Hautier Gian-Marco Rignanese 

作者机构：UCLouvainInstitute of Condensed Matter and Nanosciences(IMCN)Louvain-la-NeuveBelgium Dartmouth CollegeThayer School of EngineeringHanoverNHUSA 

出 版 物：《npj Computational Materials》 (计算材料学（英文）)

年 卷 期：2021年第7卷第1期

页      面：731-738页

核心收录：

学科分类：081203[工学-计算机应用技术] 08[工学] 0835[工学-软件工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：The authors acknowledge useful discussions and help from M.L.Evans about the MODNet development and from R.Ouyang and L.Ghiringhelli about the SISSO framework.P.-P.D.B.and G.-M.R.are grateful to the FRS-FNRS for financial support.Computational resources have been provided by the supercomputing facilities of the Universitécatholique de Louvain(CISM/UCL)and the Consortium desÉquipements de Calcul Intensif en Fédération Wallonie Bruxelles(CÉCI)funded by the Fond de la Recherche Scientifique de Belgique(FRS-FNRS)under convention 2.5020.11 and by the Walloon Region.G.H.acknowledges funding by the U.S.Department of Energy Office of Science Office of Basic Energy Sciences Materials Sciences and Engineering Division under Contract DE-AC02-05-CH11231:Materials Project program KC23MP 

主　　题：learning prediction property 

摘      要：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 *** this work,MODNet,an all-round framework,is presented which relies on a feedforward neural network,the selection of physically meaningful features,and when applicable,*** to being faster in terms of training time,this approach is shown to outperform current graph-network models on small *** 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 ***,the selection algorithm highlights the most important features and thus helps to understand the underlying physics.