Comprehensive Performance Analysis of Classifiers in Diagnosis of Epilepsy

Epilepsy becomes one of the most frequently arising brain disorder, and it is marked by unexpected occurrence frequent seizures. In this study, University Boon Database with ictal seizure disorder diagnosis epilepsy classified making use expectation maximization features as dimensionality reduction technique followed nonlinear model, namely, Gaussian mixture logistic regression, firefly algorithm, hybrid model such cuckoo search algorithm which are classifiers used for from electroencephalogram signals. The performance analyzed based on index, sensitivity, specificity, accuracy, mean square error, good detection rate, error rate. promising outcome in work indicates applied Cuckoo outperforms classification accuracy 92.19%, index 81.43%, rate 83.48%, low 15.62%, among other classifiers.