An Extended K Nearest Neighbors-Based Classifier for Epilepsy Diagnosis

In the diagnosis of epileptic seizures, classification is an important step that directly affects results. Visual inspection Electroencephalogram (EEG) a relatively common analytic method epilepsy, but it costly, time-consuming and relies on experiences doctor. Therefore, development efficient accurate seizure automatic system suitable for clinical has become urgent task. order to better solve problem early bring timely treatment patients, comprehensive representation k nearest neighbors multi-distance decision making (CRMKNN) proposed in this research. scheme, Euclidean distance Hassanat are firstly used select neighbors. Subsequently, similarity obtained through linear neighbors, calculate distribution category get discrete distance. Finally, based determine query EEG signal. verify method, we signals from Bonn university public database conducted experiments six kinds combinations. Experimental results showed our could automatically detect all situations with accuracy not less than 99.50%. At same time, compared existing methods, more effective.