Influence of Linear Features in Nonlinear Electroencephalography (EEG) Signals

Epilepsy is a critical brain disorder which can be detected through the signals captured from the brain. Electroencephalography is an efficient method used to capture signals from the brain. K nearest neighbor is one of the simplest methods used for classifying epilepsy patterns. Classification of the epilepsy signal from normal pattern will be primarily based on features extracted from brain signals. This paper discusses statistical based linear feature extraction methods such as Root Mean Square, Variance and Linear Prediction Coefficient. This paper also focuses influence of decision rules such as consensus and majority rule in the classification of epilepsy data set. Results show better classification with respect to increased k value.