Table 2 Model performance on the internal validation set.

From: Development of a machine learning-based prediction model for sepsis-associated delirium in the intensive care unit

Model

AUROC

Sensitivity

Specificity

PPV

NPV

Accuracy

Kappa

LR

0.758

0.8544

0.5058

0.7449

0.6729

0.7248

0.3795

SVM

0.780

0.8537

0.5463

0.7607

0.6886

0.7394

0.4176

RF

0.791

0.8530

0.5567

0.7647

0.6915

0.7428

0.4267

XGBoost

0.793

0.8515

0.5684

0.7692

0.6939

0.7462

0.4360

NB

0.710

0.8145

0.4476

0.7135

0.5882

0.6781

0.2755

DT

0.692

0.8004

0.5616

0.7551

0.6248

0.7116

0.3696

KNN

0.707

0.9220

0.2857

0.6856

0.6843

0.6854

0.2357

Back to article page