Dự báo khả năng bỏ học của sinh viên: Áp dụng mô hình hồi quy binary logistic

Abstract

Today, the student dropout problem is an alarming. This study aims to build a model to predict the dropout probability of in- service students - A case study at the Permanent Training Center in Long An University of Economics and Industry. The collected data includes 250 students of working and studying system who were matriculated in the academic year 2017-2018. In the study, using Binary Logistic regression for analysis, the results show that there are 5 factors affecting the student's to dropout ability, namely (1) It is necessary to design a flexible schedule so that students can choose the most suitable timetable for themselves, (2) The institute develops regulations so that it is most convenient for students to reserve their results when returning to school after the pause time, (3) The institute should develop more compact and practical way (4) Communication makes a good impression on the public, (5) Good study service helps students to stick with the college.