Epileptic seizures identification with autoregressive model and firefly optimization based classification

摘要: Identifying epilepsy cases and epileptic seizures from electroencephalogram (EEG) signals is a challenging issue, which usually needs high level of skilled neurophysiologists. Numerous works have attempted to develop tools that can provide an assistant neurophysiologist in analyzing the EEG for detection. This paper proposes new automatic framework identify classify seizure using machine learning method. In particular, feature extraction process proposed scheme utilizes autoregressive model (AR) firefly optimization (FA) procure optimal order (P). Namely, main aim FA find best (P) with minimum residual variance Akaike information criterion (AIC) as objective function algorithm. A support vector (SVM) classifier employed classification signals. The presented also effective short segment owing use AR features stage. Experiments publicly available Bonn database composed healthy (nonepileptic), interictal ictal samples show promising results accuracy.