Skip to main content

Advertisement

Table 7 AUC comparison between hybrid and expert models

From: Multi-method approach to wellness predictive modeling

Classifier Mean AUC p value
Hybrid Expert
*Naïve Bayes 0.646 0.625 <0.001
*BayesNet 0.674 0.632 <0.001
*C4.5 (J48) 0.603 0.556 <0.001
*RBF network 0.656 0.626 <0.001
*SVM (SMO) 0.695 0.652 <0.001
*Multilayer perceptron 0.700 0.664 <0.001
*Random forest 0.726 0.677 <0.001
*Bagging 0.690 0.643 <0.001
AdaBoostM1 0.554 0.550 0.499
  1. * Statistical significance, p < 0.05