International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 06 Issue: 11 | Nov 2019 www.irjet.net p-ISSN: 2395-0072 © 2019, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 125 ELECTROENCPHALOGRAM SIGNALS CLASSIFICATION USING GRADIENT BOOST ALGORITHM AND SUPPORT VECTOR MACHINE Ajao, T.A 1 , Oyewole, A.O 2 , Ojo, O.S 3 , Amore, T.O 4 , Amusan, D.G 5 and Olabode, A.O 6 1,2,4 Researcher, Federal Institute of Industrial Research Oshodi (FIIRO), Nigeria. 3,6 Research Scholar, Ladoke Akintola University of Technology, Ogbomoso, Nigeria. 5 E-Tutor, LAUTECH Open and Distance Learning Center, Ogbomoso, Nigeria. -----------------------------------------------------------------***----------------------------------------------------------------- Abstract - Automatic diagnosis of epilepsy seizure from Electroencephalogram (EEG) has been an active research in the field of biomedical science. A significant amount of classification of Electroencephalogram (EEG) signal have been proposed in recent researches; most of which achieved a very promising performance but are characterized by high false positive rate and limited by being computationally intensive. This research carried out a comparative analysis of the performance evaluation of Extreme Gradient Boost Algorithm and Support Vector Machine for the classification of epileptic seizures in human electroencephalogram (EEG). This research revealed that the XGBoost Algorithm outperformed SVM model in the classification of an EEG signal. Keywords- Extreme Gradient Boost, Support Vector Machine, Electroencephalogram (EEG) 1. INTRODUCTION In recent time, a lot of effort has been directed toward the application of computer analysis of the bio-electric signals of the human system. Several ill health conditions in man can be detected from the evaluation of the electrical signals with the body, some of the important bio-electric signals in the human system include those responsible for the heartbeat, brain signal and those in the central nervous system. However, the breakthrough in soft computing and artificial intelligence has improved the development of more effective classification, diagnostic techniques and improvements in treatment methodologies (Tzallas, et al., 2012). Soft computing technique has helped in extracting and classifying bio-signal such as Electromyography (EMG), electroencephalogram (EEG), Electrooculography (EOG) and electrocardiogram (ECG) in order to detect or treat the ailment. Different methods and techniques have been developed for detecting and classifying the electroencephalogram (EEG) as either normal or epileptic. However, complete visual analysis of EEG signal is very difficult, hence, automated means of detection is essential. Epilepsy is characterized by sudden recurrent and transient disturbances of perception or behaviour resulting from excessive synchronization of cortical neuronal networks. Epileptic seizures are divided by their clinical manifestation into partial or focal, generalized, unilateral and unclassified seizures (Tzallas, Tsipouras, and Fotiadis, 2009). The use of classification systems in medical diagnosis has increased significantly. There is no doubt that evaluation of data taken from patients and decisions of experts are the most important factors in diagnosis. Classification systems help to minimize possible errors that can be done because of a fatigued or inexperienced physician. Automated diagnostic systems have been applied to a variety of medical data, such as electrocardiograms (ECGs), electromyograms (EMGs), electroencephalograms (EEGs), ultrasound signals/images, X-rays, and computed tomographic images (AlZubi, Islam and Abbod, (2011). This research focuses on the comparative analysis of extreme gradient boost and support vector machine in the detection and classification of an EEG as either epilepsy seizure or non-epilepsy seizure for effective management of a patient suffering from epilepsy seizure. Subsequently, the rest of this paper is organized in the following sections: some reviews on related EEG signals, methodology of a Xboost and SVM, followed by results and discussion. The final section concludes the paper along with some recommendations for future research. 2. REVIEWS ON ELECTROENCEPHALOGRAM In recent time, a lot of effort has been directed toward the application of computer analysis of the bio-electric signals of the human system. Several methods for the brain function analysis such as megnetoencephalography (MEG), functional magnetic