EEG Features Extraction from Bands Delta, Theta, Alpha, Beta and Gamma.
Description
The hypothesis that will allow us to show our data is to classify the EEG signals related to real movements of the left hand, right hand, both hands, both feet, and relaxation. Each of the different files shows different features extracted from the various frequency bands composed of the EEG signals. The extracted features are pitch, level, and statistical measurements from the Delta, Theta, Alpha, Beta, and Gamma bands. Each of the samples presented in the different data sets is labeled with their respective classes to facilitate the training of the machine learning algorithms
Files
Steps to reproduce
The EEG data selected for the analysis comes from a public database. This dataset is composed of 64 EEG of 109 subjets. The chosen subjects to generate our dataset comes form 70 subject and the electrodes positioned in the motor cortex under the international 10-10 system were selected. The selected EEG signals were segmented, these segments were averaged, and an AC filter was applied to remove the noise. Then a bandpass filter is used to extract a particular band. With the EEG rhythm of interest ready, different measurements were performed: tone, level, and statistics. The whole process up to the extraction of characteristics was carried out through LabVIEW software.