DETEKSI EPILEPSI DENGAN PCA

Abstract

The main purpose of this study is to early detection of symptoms of epilepsy symptoms on the introduction of normal EEG signaling patterns with epilepsy (abnormal) EEG signals. There are 5 characteristics of statistics used are mean, variant, kurtosis, entropy, skweness. Electrodes used in EEGs usually have 19 channels: FP1, FP2, F7, F3, F2, F4, F8, C3, CZ, C4, P3, P4, PZ, O1 and OZ. While in this research only use FP1 electrode with 2 second signal cutting. Extraction of normal wave characteristics and epilepsy using PCA (principle componen analysis). PCA method is very appropriate to use if the existing data
has a large number of variables and has a correlation between variables such as EEG signals.
 The calculation of the principal component analysis is based on the calculation of eigenvalues and eigenvectors
expressing the dissemination of data from a dataset and capable of reducing the high dimension to a low dimension, without losing the information contained in the original data.
Keywords-epilepsy, EEG, FP1