Machine computer interfaces (MCI) are assistive technologies enabling paralyzed peoples to control and communicate their environments. This study aims to discover and represents a new approach on MCI using left/right motions of voluntary jaw movements stored in electroencephalogram (EEG). It extracts brain electrical activities on EEG produced by voluntary jaw movements and converts these activities to machine control commands. Jaw-operated machine computer interface is a new way of MCI entitled as jaw machine interface (JMI) provides new functionality for paralyzed people to assist available environmental devices using their jaw motions. In this article, root mean square (RMS) and standard deviation (STD) features of signals are extracted and hemispherical pattern changes are computed and compared as offline analysis approach. A statistical algorithm, principle component analysis (PCA), is used to reduce high dimensional data and two types of machine learning algorithms which are linear discriminant analysis (LDA) and support vector machine (SVM) incorporating k-fold cross validation technique are employed to identify pattern changes by utilizing the features of horizontal jaw movements stored in EEG.