Implementation bayes network based for classification study of completion

Abstract

Many factors influence the completion of the study. However, it is likely that some variables are dependent on each other. Method Bayes Network is used because it can illustrate the cause-effect relationship between completion of study with the factors that allegedly influenced as a graphical model representation so that more easily understood. With empirical data, the presence or absence of relationships among factors is calculated using mutual information. If the structure of the network has been established then subsequently calculated the probability of relationship for each event. The probability value obtained is shown to obtain a complete causality diagram. The results of the research on the classification process using student data in 2012 with Bayes Network method resulted in an accuracy of 71.09%.