A Comparison of the Learning Outcomes of Students Entered in a Private University via Multiple Entrance Programmes using Data Mining Techniques

• Main Authors: LU, YU-CHIEN, 盧囿阡
• Other Authors: YEH, SHU-YING
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/24020553673362241169

碩士 === 真理大學 === 財務與精算學系碩士班 === 104 === In recent years, the number of universities in Taiwan has been growing rapidly. The number of universities has been increasing from 123 to 158 from academic year 1991. However, an opposite trend appeared in the population growth in Taiwan which resulted in insufficient number of university students. Under the circumstances that the demographic structure is changing and the number of universities has been quickly growing, the learning outcome of students has become an important issue to all the universities. In this study, we used the grades of enrolled students from academic year 2011 to 2014 as a research material with data mining to compare the learning effects of the students enrolling universities via the different entrance programmes. The study found that under the circumstances that the places of the university admission by recommendation are opening, the number of students enrolling universities via recommendation academic year 2014 were less than 2013 in the studied private university. The reason was probably that students had not been accepted by the ideal universities and decided to try the other entrance programme. Furthermore, with CHAID decision tree algorithm we also learnt that the semester grades of the subject programming of the students enrolling the studied university via recommendation might be able to be predicted by their English grades in their college entrance exam in subject competence. The apriori algorithm of association rules were also used to find out that most students achieved more than 80 points in calculus in the Department of Public Finance and Taxation, in which the lift value was 2.68. The decision tree algorithm was also applied to data of the students enrolling university via college entrance exam in specified subjects. The semester grades of accounting were learnt to be able to be predicted by the scores of chemistry, citizen and society, mathematics in the college entrance examination in specified subjects. In addition, most students also achieved more than 80 points in calculus in the Department of Public Finance and Taxation, in which the lift value was 1.9 in the apriori algorithm. It shows that there was a high probability that the teacher of calculus in the Department of Public Finance and Taxation mark grades more generously than any other departments.