EEG Signal Classification Using Ensemble Empirical Mode Decomposition

Abstract: A method for feature extraction from electroencephalogram (EEG) signals using ensemble empirical mode decomposition (EEMD) is developed. Its use is motivated by the fact that the EEMD gives an effective time-frequency analysis of non-stationary signals. The existing work makes use of EMD which involves in taking third order IMFs and also mode mixing is the one of the major problem in EMD. The proposed method overcomes the problem of mode mixing by applying a white noise to the signal on decomposition. Five different datasets are collected and used for analysis. The result of EEMD is the intrinsic mode functions which give the decomposition of a signal according to its frequency components. Temporal moments, and spectral features including spectral centroid, coefficient of variation and the spectral skew of the IMF is used for feature extraction from EEG signals. The calculated features are fed into the standard support vector machine (SVM) for classification purposes.

Keywords: EEMD, EMD, Support vector machine, Temporal and Spectral features.