Detection of Epileptic Seizure from EEG Signal Using Discrete Wavelet Transform and J48 Classifier

Detection of Epileptic Seizure from EEG Signal Using Discrete Wavelet Transform and J48 Classifier

Abstract— The application on wavelet decomposition and J48 Decision tree model for Electroencephalograph signals (ECG) and its classification. Decision making perform by feature extraction using the discrete wavelet transform < Final Year Projects 2016 > DWT with different function and the EEG classification using J48 classifiers. Brain Computer interfaces (BCI) are technologies that control communication among the human brain and peripheral devices. Headways in psychological neuroscience and mind imaging advances have begun to present us with the ability to interface specifically with the human cerebrum.This is made conceivable through
the utilization of sensors that can screen a percentage of the physical procedures that happen inside of the
cerebrum that relate with specific types of thought in these frameworks, clients expressly control their mind action rather than utilizing engine developments to create signals that can be utilized to control PCs In this work, the extracted features using DWT are fed into J48 classifier for seizure detection of EEG signal with an accuracy of 95%. The proposed J48 model achieved higher accuracy rates than that of the K-NN algorithm.