Prediction model of band-gap for AX binary compounds by combination of density functional theory calculations and machine learning techniques

摘要

Machine learning techniques are applied to make prediction models of the G0W0band-gaps for 156 AX binary compounds using Kohn-Sham band-gaps and otherfundamental information of constituent elements and crystal structure aspredictors. Ordinary least square regression (OLSR), least absolute shrinkageand selection operator (LASSO) and non-linear support vector regression (SVR)methods are applied with several levels of predictor sets. When the Kohn-Shamband-gap by GGA (PBE) or modified Becke-Johnson (mBJ) is used as a singlepredictor, OLSR model predicts the G0W0 band-gap of a randomly selected testdata with the root mean square error (RMSE) of 0.54 eV. When Kohn-Sham band gapby PBE and mBJ methods are used together with a set of various forms ofpredictors representing constituent elements and crystal structures, RMSEdecreases significantly. The best model by SVR yields the RMSE of 0.18 eV. Alarge set of band-gaps estimated in this way should be useful as predictors formaterials exploration.