Epileptic seizure classification of electroencephalogram signals using extreme gradient boosting classifier

Epilepsy causes repeated seizures in an individual's life, which transient irregularities the brain's electrical activity. It results different physical symptoms that are abnormal. Various antiepileptic drugs fail to minimize patient seizures. The electroencephalogram (EEG) signal recordings provide us with time-series data set for epileptic seizure detection and analysis. These signals highly nonlinear inconsistent, they recorded over time. Predicting ictal period (seizure at time of epilepsy) is thus a challenging task naked eye medical practitioners. machine learning techniques applied identify seizure's occurrence its classification multiple domains. A model based on extreme gradient boosting (SCLXGB) proposed here EEG signals. SCLXGB implements binary benchmark dataset. Compared K-nearest neighbor, linear regression, Decision treebased models, achieves best area under receiver operating curve (AUC) 0.9462 accuracy 96% signifies accurate prediction non period. was validated by taking performance metrics indicate non-occurrence patients more appropriately.