Electroencephalography signals are obtained by measuring the electrical current in the brain. The classification of these signals are especially important, as they contribute to the diagnosis, and treatment of disorders related to brain signals. In this study, the data set of the University of Bonn, which is the most widely used data set for the diagnosis of epilepsy, was used in this field. In order to obtain meaningful results from this data set consisting of signals from five different subjects, firstly, data filtering, feature extraction and feature selection methods have been used first. Later, these methods were then compared in terms of their contribution to classification success. First, statistical properties were extracted from the filtered data by the Discrete Wavelet Transform method, and then the subset of the features that gave the best classification result was selected using the Differential Evolution Algorithm. The classification success of the data set with the selected features has been tested with the Support Vector Machines. With the method used, better results were obtained than similar studies in separating some classes. In the classification of some double and triple sets, accuracy rates of 0.98 and 0.94, respectively, were obtained.