I. Introduction
Approximately 1% of the world population suffers from epilepsy [1]. Epilepsy is a common chronic neurological disorder that may cause brief electrical disturbance in the brain and producing to uncontrollable movement and trembling [2]. Electroencephalogram (EEG) is a record of electrical activity of the brain. Therefore, EEG signal can be used to analysis the behavior of the brain and detect epilepsy. The EEG recording is one of the most important tools for the diagnosis of neurological diseases [3]. Feature extraction based on Fourier transform (FT) can perfectly isolate the frequency content of a signal, but cannot localize when the components happened in time [4]. In general, since the EEG signal is non-stationary, it is most suitable to utilize the time-frequency domain transforms like discrete wavelet transform (DWT). Therefore wavelet transform is a powerful tool for feature extraction of EEG signals [5]–[7].