Machine learning prediction of magnetic properties of Fe-based metallic glasses considering glass forming ability

作     者：Xin Li Guangcun Shan C.H.Shek Xin Li;Guangcun Shan;C.H. Shek

作者机构：School of Instrumentation Science and Opto-electronics EngineeringBeihang UniversityBeijing 100191China Department of Materials Science and EngineeringCity University of Hong KongKowloon TongHong Kong SARChina 

出 版 物：《Journal of Materials Science & Technology》 (材料科学技术（英文版）)

年 卷 期：2022年第103卷第8期

页      面：113-120页

核心收录：

学科分类：08[工学] 081104[工学-模式识别与智能系统] 0817[工学-化学工程与技术] 0806[工学-冶金工程] 0805[工学-材料科学与工程（可授工学、理学学位）] 080502[工学-材料学] 0703[理学-化学] 0802[工学-机械工程] 0811[工学-控制科学与工程] 0801[工学-力学（可授工学、理学学位）] 0702[理学-物理学] 

基　　金：financially supported by National Natural Science Foundation of China(No.21771017) the Fundamental Research Funds for the Central Universities。 

主　　题：Metallic glasses Soft magnetic properties Glass forming ability Machine learning Non-linear regression 

摘      要：Fe-based metallic glasses(MGs)have shown great commercial values due to their excellent soft magnetic properties.Magnetism prediction with consideration of glass forming ability(GFA)is of great signifi-cance for developing novel functional Fe-based MGs.However,theories or models established based on condensed matter physics exhibit limited accuracy and some exceptions.In this work,based on 618 Fe-based MGs samples collected from published works,machine learning(ML)models were well trained to predict saturated magnetization(B_(s))of Fe-based MGs.GFA was treated as a feature using the experimental data of the supercooled liquid region(△T_(x)).Three ML algorithms,namely eXtreme gradient boosting(XGBoost),artificial neural networks(ANN)and random forest(RF),were studied.Through feature selection and hyperparameter tuning,XGBoost showed the best predictive performance on the randomly split test dataset with determination coefficient(R^(2))of 0.942,mean absolute percent error(MAPE)of 5.563%,and root mean squared error(RMSE)of 0.078 T.A variety of feature importance rankings derived by XGBoost models showed that T_(x) played an important role in the predictive performance of the models.This work showed the proposed ML method can simultaneously aggregate GFA and other features in ther-modynamics,kinetics and structures to predict the magnetic properties of Fe-based MGs with excellent accuracy.