Research Article

A Novel Approach for Sleep Arousal Disorder Detection Based on the Interaction of Physiological Signals and Metaheuristic Learning

Table 1

The error assessment of sleep steps classification for a two-class problem using EEG signals.

RepetitionsSVMSVM (RBF)SVM (RBF) + GASVM (RBF) + PSOSVM (RBF) + DSOSVM (RBF) + mDSO
MinMaxMinMaxMinMaxMinMaxMinMaxMinMax

10-fold (1)0.110.150.080.110.080.120.050.080.030.050.020.04
10-fold (2)0.090.170.080.120.090.120.060.090.040.080.020.05
10-fold (3)0.150.180.090.130.080.120.060.080.040.060.010.05
10-fold (4)0.110.160.080.120.060.130.060.080.020.070.030.05
10-fold (5)0.140.170.110.130.110.120.080.080.040.080.020.04
10-fold (6)0.130.180.100.140.100.110.090.090.040.060.010.03
10-fold (7)0.110.160.100.140.090.110.060.110.030.060.010.03
10-fold (8)0.090.160.070.140.070.090.050.080.050.070.020.03
10-fold (9)0.100.140.120.130.110.110.090.070.050.090.030.05
10-fold (10)0.120.150.100.130.090.090.070.120.040.080.020.04
Av0.1150.1620.0930.1290.0880.1120.0670.0880.0380.0710.0190.041

Bold values are the best values obtained.