Abstract: This paper aims to classify three mental task which includes left hand movement imagination, right hand movement imagination and word generation. The classification of three or more tasks is a crucial process. It depends upon how the features are extracted and classified. Various methods are available for extracting features from raw electroencephalographic(EEG) data, but the utilization of irrelevant or superfluous features affects the performance of the classifier. Thus extracting and selecting features is a very crucial step for proper classification. This study demonstrates GA as a feature selection method. Here EEG data from BCI III dataset V is used in which features are extracted by using PSD. Then features are selected by using Genetic Algorithm and finally they are fed to the classifier for classification. For classification Neural Network (NN) is used.

Keywords: EEG, Genetic Algorithm ,Mental task, Neural Network, PSD.