To search, Click below search items.


All Published Papers Search Service


Impact of Sampling Features on EEG Classification


Ala Alam Falaki and Seyyedeh Hoora Fakhrmoosavy


Vol. 17  No. 11  pp. 62-67


In this paper, a sampling method is going to be introduce for classifying Electroencephalogram (as known as EEG) signals. This method consists of three steps. (i) Reduce EOG artifacts. (ii) Calculating the bandpower of the signal. (iii) Finding the best time segment of features with highest classification accuracy in the range of bandpower peaks. The Butterworth algorithm used for feature extraction and the classification accuracy measured by Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), Na?ve-Bayes (NB) and Random Forest (RF) algorithms. For the experiment, dataset 2b from BCI competition IV that recorded in 3 channels for motor imagery tasks were studied, two different mental tasks are examined for each subject in two class labels for right- and left- hand movement.


EEG, Motor Imagery, Hand Movement, band power, Sampling