EEG signal analysis using sparse approximations

Abstract

Electroencephalogram (EEG) is physiological signal generated in the brain. Electroencephalography is a method to record the electrical activity of the brain in order to detect the abnormalities of the brain. However, EEG also detects the signals which are not originated from the brain called artifacts. This paper deals with the analysis and extraction of EEG signals in sparse representation using sparse algorithms Orthogonal Matching Pursuit (OMP) and LASSO. OMP is an iterative greedy algorithm which replaces optimization problem in each step of Matching Pursuit(MP), an earlier algorithm for solving sparse approximation problems by least squares minimization. LASSO is an optimization technique which involves regression analysis concepts. Sparse approximations are used in practical applications like feature extraction, Denoising, Inpainting etc.