Table 2.

Performance metrics of 6 classifiers on the train set and test set.

Performance	Accuracy (%)	lTPR	mTNR	Precision	nAUC	Kappa	°MAE
gLR
Train set	98.70	0.980	0.993	0.987	0.996	0.973	0.02
Test set	98.72	0.987	1.000	0.988	1.000	0.974	0.01
hRC
Train set	96.40	0.967	0.961	0.964	0.997	0.928	0.07
Test set	98.72	0.974	1.000	0.988	1.000	1.000	0.05
iSMO
Train set	97.72	0.961	0.993	0.978	0.977	0.954	0.02
Test set	97.44	0.974	0.974	0.974	0.974	0.949	0.03
jRF
Train set	97.72	0.974	0.980	0.977	0.997	0.954	0.10
Test set	97.44	0.974	0.974	0.974	0.999	0.949	0.08
kNB
Train set	97.01	0.948	0.993	0.972	0.994	0.942	0.06
Test set	98.72	0.974	1.000	0.988	1.000	0.974	0.05
Kstar
Train set	96.08	0.922	1.000	0.964	0.997	0.921	0.10
Test set	96.15	0.949	0.949	0.974	0.997	0.923	0.10

^glogistic regression.

^hRandom Committee.

ⁱSequential minimal optimization.

^jRandom Forest.

^kNaive Bayes.

^lTrue Positive Rate.

^mTrue Negative Rate.

ⁿArea under curve.

^°Mean absolute error.

The best performance metrics for each set are highlighted in bold.