Neural network based seizure detection system using statistical package analysis

Due to the unpredictable interruptions within functions of human brain, disturbance occurs and it affects behavior is equally laid low with frequent occurrence termed as seizures. Therefore, proposed system detects seizure using machine learning algorithms. The electroencephalogram (EEG) contains information brain detect seizure. objective evaluate performance classifiers K-nearest neighbors (KNN), artificial neural network (ANN), support vector (SVM) principal component analysis (PCA) by comparing accuracy classifier. This work uses total 11,500 EEG samples from UCI repository. detection was done in two ways. First method, features extracted signal classification techniques are classify second method algorithm improve significant selections dataset. outcomes analyzed statistical package for social science (SPSS) tools. ANN achieved 96% efficiency (p less than 0.05) comparison different classifiers. It would be prudent conclude that demonstrated best accuracy, sensitivity, specificity.