Student Study Timeline Prediction Model Using Naïve Bayes Based Forward Selection Feature

The student study period is one of the factors that show a student's academic performance. Universities are required to be able to keep students able to complete their studies on time so that there is no buildup of the number of students who have not graduated. Therefore, from the academic data students conducted data mining classification using naive Bayes algorithm. But because of the many attributes, to speed up this naive Bayes modeler, it is supported by the selection of the forward selection feature. In the Selection process, the feature generates 5 selected attributes that affect the dataset. While from this classification process obtained the accuracy value of the prediction model naive Bayes increased from 90.00% to 92.94% after adding a forward selection feature. With this high accuracy score, prediction models can be applied in policymaking to prevent students from graduating on time.