Skip to main content

Table 4 Performance comparisons of our stacking model and the eight candidate models

From: A stacking-based model for predicting 30-day all-cause hospital readmissions of patients with acute myocardial infarction

Model

AUC

Accuracy

Sensitivity

Specificity

DT

0.681 ± 0.02

0.768 ± 0.03

0.487 ± 0.06

0.810 ± 0.04

SVM

0.707 ± 0.03

0.765 ± 0.01

0.480 ± 0.03

0.808 ± 0.01

RF

0.701 ± 0.02

0.768 ± 0.01

0.502 ± 0.05

0.807 ± 0.01

ET

0.709 ± 0.02

0.760 ± 0.02

0.500 ± 0.03

0.798 ± 0.02

GBDT

0.710 ± 0.02

0.764 ± 0.02

0.501 ± 0.04

0.803 ± 0.02

ADB

0.702 ± 0.02

0.769 ± 0.03

0.502 ± 0.03

0.809 ± 0.03

Bagging

0.704 ± 0.02

0.769 ± 0.01

0.512 ± 0.03

0.808 ± 0.01

XGB

0.713 ± 0.02

0.768 ± 0.02

0.513 ± 0.03

0.806 ± 0.02

Stacking Model

0.720 ± 0.02

0.772 ± 0.01

0.515 ± 0.04

0.810 ± 0.01

  1. Font bold: the optimal values. DT Decision tree, SVM Support vector machine, RF Random forest, ET Extra trees, GBDT Gradient boosting decision tree, ADB AdaBoost, Bagging Bootstrap aggregating, XGB Extreme gradient boosting