Analysis and classification of electroencephalography signals

摘要: EEG signal processing is one of the hottest areas research in digital applications and biomedical research. Analysis signals provides a crucial tool for diagnosis neurobiological diseases. The problem classification into healthy pathological cases primarily pattern recognition using extracted features. Many methods feature extraction have been applied to extract relevant characteristics from given data. data was collected publicly available source. Three types were classified viz. recorded volunteers having their eyes open, epilepsy patients epileptogenic zone during seizure-free interval, epileptic seizures. done by computing discrete wavelet transform spectral analysis AR model. coefficients compress number points few Various statistics used further reduce dimensionality. obtained burg auto-regressive method provide important features signals. Classification committee neural network robust improved performance over individual members committee. F-ratio based dimension reduction technique without affecting accuracy much.