Classifying electroencephalogram signals using an innovative and effective machine learning method based on chaotic elephant herding optimum

Abstract

The field of electroencephalography (EEG) has made significant contributions to our understanding of the brain, our understanding of neurological diseases, and our ability to treat such diseases. Epileptic seizures, strokes, and even death can all be detected with the use of the electroencephalogram, a diagnostic technique used to record electrical activity in the brain. This research suggests using binary classification for automated epilepsy diagnosis. Patients' EEG signals are pre-processed after being recorded. On the basis of the results of the feature extraction technique, the best traits are picked for further examination by means of a structured genetic algorithm. The EEG data are analysed and categorized as either seizure-free or epileptic seizure-related based on the assumption of feature optimization utilizing the support vector classifier. As a result, categorizing EEG signals is an ideal application for the suggested technique. For this purpose of accelerating the implementation of distributed computing, a CEHOC (Chaotic Elephant Herding Optimization based Classification) is used to classify the vast scope of various datasets. The results show that the CEHOC algorithm is more effective than previous versions. Precision, recall, F score, sensitivity, specificity, and accuracy are some of the metrics used to assess the effectiveness of the work provided here. The suggested work has a 99.3019% accuracy rate, a 98.2018% sensitivity rate, and a 99.1125% specificity rate. There was an F score of 99.3204%, a precision of 99.1019%, and a recall of 98.3015%. These numbers indicate that the planned action was successful.