Research Article

Stress Classification by Multimodal Physiological Signals Using Variational Mode Decomposition and Machine Learning

Table 2

Classification results for the percentage split.

SetsStatistical parametersMPLNSVMMPLNSVMMPLNSVM
Modes 1-4Modes 5-8Modes 1-8

Good performer vs. bad performerSensitivity97.261.239868.679870.65
Specificity96.365.129868.679870.65
Accuracy97.261.239868.679870.65
F-measure0.950.650.970.5710.69
Kappa statistics0.940.630.980.560.990.69

Good performer vs. bad performer (female)Sensitivity78.5662.3481.6768.6783.7869.67
Specificity78.962.3782.2368.6783.6769.67
Accuracy76.5662.7881.7868.6783.6769.67
F-measure0.670.640.820.650.870.67
Kappa statistics0.680.630.820.650.880.69

Good performer vs. bad performer (male)Sensitivity79.5663.4380.2168.6782.6770.02
Specificity78.7863.5680.1268.6782.5670.02
Accuracy78.6763.5680.3268.6782.6770.02
F-measure0.870.650.790.680.830.69
Kappa statistics0.870.650.790.670.830.69

Good performer vs. bad performer (frontal region)Sensitivity9870.7898.4570.679972.67
Specificity9878.7898.4570.679972.67
Accuracy9867.8998.4570.679972.67
F-measure0.980.740.990.6910.71
Kappa statistics10.7810.710.71

Good performer vs. bad performer (temporal region)Sensitivity99.867.7899.9961.2399.9975.56
Specificity97.7867.7899.9965.1299.9975.56
Accuracy99.867.7899.9961.2399.9975.56
F-Measure0.990.740.990.5910.72
Kappa statistics10.7410.5910.76

Good performer vs. bad performer (occipital region)Sensitivity78.566080.2362.3489.3476.7
Specificity78.96080.3462.3789.4376.7
Accuracy76.566080.3262.7889.3476.7
F-measure0.880.580.890.620.850.74
Kappa statistics0.850.580.860.620.850.74

Good performer vs. bad performer (central region)Sensitivity75.7864.6779.7863.4380.5668.78
Specificity76.8964.6779.5663.5680.5668.78
Accuracy78.676479.3463.5680.5668.78
F-measure0.760.670.750.620.830.56
Kappa statistics0.780.650.780.620.830.56