Automatic Identification of Epileptic Seizure Using Kelm Optimized By Grey Wolf Algorithm

Epilepsy is the common neurological disorder where nerve cell activity is disturbed which results in causing seizures. Automatic epileptic seizure detection is an essential requirement for clinical diagnosis of epilepsy. Classifying facts is the usual task in machine learning. Epileptic seizure identification using Support Vector Machine (SVM) have few limitations like its outcomes lacks transparency. Extreme Learning machine is computationally uncomplicated and yields definite results also because of its quick learning pace, (ELM) is utilised to pick input data. The motivation of this research is to boost the training rate and correctness of the ELM. Kernel Extreme Learning Machine (KELM) is chosen for refining generalization capacity and to improve the classification accuracy nature inspired swarm intelligence Grey Wolf Algorithm (GWO) is adapted. The grey wolf algorithm optimizes the KELM parameters. The GWO-KELM classifier is laid on Epileptic Seizure Data Set from UCI and the experimental results such as learning accuracy, learning error and classification accuracy are compared with traditional classifiers. By performing analysis with the proposed GWO- KELM classifier faster learning speed, accuracy with high precision and low error rate is achieved and proposed classifier outperforms other classifiers in identification of epileptic seizures.