Research Article
Selection of Neural Oscillatory Features for Human Stress Classification with Single Channel EEG Headset
Table 2
A comparison of the performance parameters of algorithm used for classification of human stress using single channel headset.
| | Neural Oscillations | ACC (%) | Kappa | F-Measure | Time (sec) | RMSE | RAE | Classifier |
| | All Oscillations ([23]) | 71.43 | 0.21 | 0.531 | 0.080 | 64.61 | 114.03 | Naive Bayes | | Low Beta ([24]) | 71.43 | 0.2 | 0.662 | 0.004 | 52.00 | 65.29 | SVM | | Discrete Cosine Transform ([25]) | 72.00 | − | − | − | − | − | K-nearest Neighbour | | delta, theta, alpha and beta and IMFs ([20]) | 83.33 | − | − | − | − | − | Naive Bayes | | Low Beta, High Beta, Low Gamma (Proposed) | 78.57 | 0.44 | 0.662 | 0.010 | 48.46 | 98.75 | SVM |
|
|
