A Statistical Based Approach for Feature Extraction from EEG Signals
Brain Computer Interface is one of the most essential component of the intelligence based computing approach that acquires brain signals, analyze and translate into commands. Feature Extraction technique is used to improve the classification efficiency. In this work various machine intelligence based feature selection technique were used which helps to maximize relevance and minimize redundancy. In this paper EEG signals are preprocessed and features are extracted using statistical methods and the features are selected using InfoGainAttributeEval, OneRAttributeEval, ReliefFAttributeEval. The selected features are classified by the classifier Random Forest. Experimental results shown that the feature selection method InfoGainAttributeEval gives better result compared to other state of art methods.