Table 1 The comparison of prediction performance of Eg models with different features and ML regression algorithms by three evaluation metrics (R2, RMSE and MAE). The best results are highlighted in bold.

From: Data-driven design of high-performance MASnxPb1-xI3 perovskite materials by machine learning and experimental realization

Method

14 features

One feature

 

R2

RMSE

MAE

R2

RMSE

MAE

LR

0.6864

0.0751

0.0630

0.6064

0.0841

0.0707

SVR

0.8775

0.0469

0.0422

0.8221

0.0565

0.0516

KNR

0.8997

0.0425

0.0340

0.8833

0.0458

0.0365

RFR

0.9105

0.0401

0.0284

0.9065

0.0410

0.0344

GBR

0.9172

0.0386

0.0325

0.9072

0.0408

0.0283