SEIZURE EEG SIGNALS DETECTION AND CLASSIFICATION USING WAVELET TRANSFORM AND WOA BASED LLRBFNN MODEL

This paper presents a novel classification WOA-LLRBFNN (Whale optimization algorithm- Local Linear Radial Basis Function neural network) Model for classification of seizure EEG signal. The Bonn dataset has been applied for classification and detection of the seizure. The wavelet transform has been used for detection of seizure from the dataset. There are three wavelets such as Daubechies, coiflet and symlet are considered for detection and comparison results are presented. It is found that coiflet transform has shown good detection results in comparison to the other wavelets. Further, the proposed WOA based LLRBFNN accuracy results are compared with LLRBFNN, LLWNN (Local linear wavelet transform), RBFNN model to show the robustness of the model. It is found that the WOA-LLRBFNN model achieved an accuracy of 99.12% which is higher than the other models.