Batman Universitesi Yasam Bilimleri Dergisi
Cilt 1, Sayı 2  Ocak-Haziran 2012  (ISSN: 2147-4877, E-ISSN: 2459-0614)
Yılmaz KAYA; Ö.Faruk ERTUĞRUL; Ramazan TEKİN

NO Makale Adı
1356123443 Epileptic EEG Detection Using Classifiers Decision Trees and Decision Rules

Epilepsy is a well defined chronic neurological disorder usually characterized
by seizures which are transient symptoms of abnormal or hyper synchronous
neuronal activity in the brain. Electroencephalography (EEG) is one of the most common
methods in diagnosis of epilepsy.In this study, decision trees( ADTree, Functional
Tree, J48, NBTree) and decision rules (Furia, DTNB, Jrip, PART, Ridor) classifier
algorithms were implemented for statistical features of EEG signal that obtained with
discreate wavelet transformation. The obtained results show that decision rules and
trees classifier algorithms performed a good performance and high prediction on epileptic
EEG signals. For various combinations of epileptic EEG data sets the accury
was ranged from 96.6% to 99.70%.