Identification of Fatal Familial Insomnia Sleep Disorder Using Ga Optimization with Svm Classifier

Abstract

Among the genetic disorders in brain Fatal familial insomnia (FFI) is one of the rare disorder. The inability of sleeping is called as FFI which leads to the process of becoming progressively worse mentally and physically. The recordings of brain functioning is known as Electroencephalogram (EEG) and which seeks a vital role in observing and finding the sleep disorders like FFI. The EEG senses the brain functionality. In this paper , genetic optimization based SVM classifier is used to identify the sleep disorder. The proposed technique is used to optimize the features which are obtained from features reduction techniques like PCA, ICA and LDA. For performing the experimental results PhysioNet database is obtained. Preprocessing of the data, performing feature reduction, next feature optimization and then classification using Support vector machine is the process for identifying the FFI. The performance measures such as accuracy, sensitivity, specificity, F1-Score, Recall are computed . The comparison of results obtained using different condition are observed and the ranges are determined for normal class, Bruxism class and FFI class.