Table 2 Performance of prognostic models built by machine learning algorithms in the training and test sets (area under the ROC curve).

From: Development of a machine learning-based model for prognostic prediction in melanoma

 

1-year survival

3-year survival

5-year survival

Train set

 RF

0.8831

0.8654

0.8566

 DT

0.7342

0.7089

0.7061

 XGBoost

0.8790

0.8634

0.8583

 CatBoost

0.8895

0.8730

0.8667

 LightGBM

0.8706

0.8562

0.8514

Test set

 RF

0.7416

0.7469

0.7426

 DT

0.6836

0.6719

0.6727

 XGBoost

0.7380

0.7621

0.7646

 CatBoost

0.7551

0.7671

0.7690

 LightGBM

0.7274

0.7506

0.7589

  1. ROC receiver operating characteristic curve; RF random forest; DT decision tree; XGBoost extreme gradient boosting; CatBoost categorical boosting; LightGBM light gradient boosting machine.
Back to article page