Development of Rules for Predicting Academic Learning Outcomes of Freshmen Students using Data Mining Techniques

Abstract

In this research investigation, the researchers apply the knowledge of data mining to analyze the academic learning outcomes of the students in various courses in the first year study plan of the Department of Statistics, Faculty of Science, King Mongkut’s Institute of Technology Ladkrabang. The researchers also establish the relationship rules between the learning outcomes of the courses taken with a grade point average (GPA) lower than 2.00 using the association rules with a priori algorithm and the decision rules for data classification with the technique of J48. The results could be used to make a plan for the study of the students. The study found that the rules used for the classification of the learning outcomes of the students under study with a GPA lower than 2.00 and those with a GPA higher than 2.00 using the J48 technique yielded the accuracy of 91 percent. The number of the association rules of the courses that affected a GPA lower than 2.00 of the students under study was 5 with the reliability at 1.00 and Lift>1.00.