INCREASING ACCURACY OF THE EPILEPSY SIGNAL CLASSIFICATION Hindarto hindarto,ade Eviyanti, M Abror
UNIVERSITAS MUHAMMADIYAH SIDOARJO
Abstract
Epilepsy is a manifestation of brain disorders with various etiologies, but with a single, characteristic symptom, namely periodic and reversible attacks, epilepsy is characterized by an excess amount of electricity coming out of brain cells, which can cause seizures and abnormal movements. EEG signals in epileptic seizures have a characteristic pattern that allows health professionals to distinguish them from normal conditions (nonseizure). Many methods are used by researchers to perform pattern recognition of epilepsy and non-epilepsy EEG signals, in this study researchers tried to use sampling techniques as a feature of extracting epilepsy signal features and the K-NN method to identify epilepsy signal patterns. The data of this study took epilepsy signal data from the University of Bonns Epileptologie clinic which consisted of data set A, open eye normal signal, set B normal closed eye signal, set C in epilepsy zone, set D enter epilepsy, set E seizure epilepsy. In this study, researchers tried to classify data set A, data for normal people and data set E, data for people who have epilepsy. Data set A consists of 100 EEG signals and data set E consists of 100 EEG signal data. 50 EEG signal data are used for training data and 50 EEG signal data for testing data. The best results with a value of K = 1 to K = 9 obtained for the classification of EEG signals using K-NN is 100% of the tested signal data.
