Table 7 ML models’ statistics and performance metrics.

From: Modeling of nitrogen solubility in normal alkanes using machine learning methods compared with cubic and PC-SAFT equations of state

Model

 

RMSE

SD

R2

k-NN

Total

0.0276

0.4632

0.9802

Train

0.0259

0.4799

0.9825

Test

0.0336

0.3901

0.9716

Random forest

Total

0.0208

0.2361

0.9886

Train

0.0170

0.1820

0.9931

Test

0.0319

0.3826

0.9760

XGBoost

Total

0.0241

0.8669

0.9859

Train

0.0219

0.9005

0.9884

Test

0.0316

0.7181

0.9767

LightGBM

Total

0.0295

0.7002

0.9790

Train

0.0276

0.6415

0.9801

Test

0.0328

0.8981

0.9729

CatBoost

Total

0.0147

0.1739

0.9943

Train

0.0125

0.1219

0.9960

Test

0.0213

0.3032

0.9887

Back to article page