Investigation of Epileptic EEG Data Using Ensemble Empirical Mode Decomposition

In this study, it was aimed to classify the epileptic and normal EEG data by using the Ensemble Empirical Mode Decomposition (EEMD) method. For this purpose, we studied with 3 data groups and 30 data from each group were examined. Firstly, data were decomposed into intrinsic mode functions (IMEs) using EEMD. Decomposer features were calculated from the 1st IMF of the EEMD expansion of EEG signals from epileptic and healthy subjects. Power density is estimated by Welch and Periodogram methods and high frequency moments are calculated. When the moment value obtained by both methods were examined, it was observed that the epileptic EEG data could be separated with high success from the normal EEG data.