A STATISTICAL FEATURE EXTRACTION IN WAVELET DOMAIN FOR MOVEMENT CLASSIFICATION: A CASE STUDY FOR EYES OPEN, EYES CLOSED, AND OPEN/CLOSED FIST TASKS

Abstract

Analysis of brain signals constitute an importance,especially for paralyzed people or people suffer from motor disabilities. Forthis aim, some studies have been evaluated to measure signals from the scalp toprovide non-muscle control arguments. Brain-Computer Interface Systems turnsthese signals into device signals that are controllable at the level ofthought. In this paper, we classify diverse tasks according to EEG(electroencephalogram) signals. Then pre-processing, feature extraction andclassification steps are hold. For classification, we use FLDA, Linear SVM,Quadratic SVM, PCA, and k-NN methods. The best result is obtained by usingk-NN.