Skip to main content

Table 9 Test statistics of the top five candidate ensemble models with optimization

From: Predicting students at risk of academic failure using ensemble model during pandemic in a distance learning system

Candidate ensemble and its

sub-models

TN

FP

FN

TP

ET + RF + LR

131

14

110

359

QDA + LR

121

24

89

380

GB + ET + RF + LR

121

24

91

378

GB + DT + LR

117

28

90

379

GB + DT + LR + ANN

118

27

91

378

  1. Italic characters show highest true negative prediction.
  2. Gradient Boosting (GB), Quadratic Discriminant Analysis (QDA), Extra Trees (ET), Decision Tree (DT), Random Forest (RF), Logistic Regression (LR), and Artificial Neural Network (ANN)