Predicting the Probability of Student's Academic Abilities and Progress with EMIR and Data from Current and Graduated Students

In 2016, Kobe Tokiwa University constructed an office for institutional research (IR) promotion. The purpose of this office is to propose, manage, arrange, and collect information on students at the university not only as a general management strategy, but also to support enrollment management. Our database currently contains 3,495 points of data (i.e., headcounts), each containing 1,246 items of numerical value. Last year, we reported on an analysis that focused on the "student dropout" phenomenon by using these data from both current graduate and dropout students. This year, we formulated a research question that is centered on predicting the probability of students' progress and academic abilities through Enrollment Management / Institutional Research (EMIR). We obtained results with these data by processing them through a machine learning technique using random forest, which yielded a correction rate of about 90%.