Eigen Vector Method with Swarm and Non Swarm Intelligence Techniques for Epileptic Seizure Classification

One of the most commonly occurring neurological disorder in the human brain is epilepsy. It is a long-term chaos in the Central Nervous System (CNS) that severely affects the life of an individual due to repeated seizures. In the electrical activity of the brain, a seizure is nothing but a slight or serious transient irregularity that tends to disturb the cortical regions of the brain and produces symptoms such as muscle spasms, sensory illusion, fatigueness, memory lapse, attention lapse etc. For the diagnosis of epilepsy, Electroencephalography (EEG) signals is used widely. In this work, Eigen vector method utilizing Pisarenko's technique is utilized to extract the features from EEG signals. Then the extracted features are optimized with two techniques, one is a swarm intelligence technique and the other is a non-swarm intelligence technique. The swarm intelligence technique used here is a Bat optimization algorithm and the non-swarm intelligence technique used here is a Biogeography based optimization algorithm. Finally, it is classified with the help of Decision Trees, Multilayer Perceptron (MLP) and Random Forest (RF) classifiers. Results show that a highest classification accuracy of 95.57% is obtained when Eigen Vector technique is utilized with Bat optimization algorithm and classified with Random Forest (RF) classifier.