Feature based classification of Electrooculogram using SVM and KNN

Abstract

In recent years, Electrooculography (EOG) is one of the technique to measure biomedical signal for analyzing the eye movement patterns in corneo-retinal which located between the front and back of the human eye. This electrographic signalused to measure the eye movements from left - right or top - bottom which creates the electrical deflections. In this paper, the EOG datasets was collected from five participants and before feature extraction the respective data are preprocessed which helps to reduce the dimensions of the given data. Then the features are extracted by SVM (Support Vector Machine), and KNN (K – Nearest Neighbor). Hence, the performance evaluation was proved that KNN gives 80.50% accuracy than SVM classifier.