Table 3. Comparisons of performance of the three different regression methods.

Obviously, the Random Forest demonstrates the best capability to correlate the SampEn with BIS. SVM and ANN exhibit similar regression effects. For Random Forest, it has the best performance in terms of all the metrics among all models.

	CC	AUC	MAE
SVM	0.76 ± 0.09	0.95 ± 0.04	9.08 ± 2.72
ANN	0.79 ± 0.08	0.96 ± 0.04	8.88 ± 2.46
Random Forest	0.90 ± 0.05	0.98 ± 0.02	5.22 ± 1.12

Notes.

Values are presented with mean ± STD. Markers symbolize the statistical significant Student t-test.

SVM

support vector machine

ANN

artificial neural network

CC

correlation coefficient

AUC

area under curve

MAE

mean absolute error