CHURN PREDICTION FOR A PRIVATE UNIVERSITY IN SOUTH TAIWAN USING MACHINE LEARNING TECHNIQUES

• Main Authors: Chun-YiPeng, 彭淳奕
• Other Authors: Lih-Chyun Shu
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/n2e9n3

碩士 === 國立成功大學 === 會計學系 === 103 === Currently, universities in Taiwan are facing many management pressures on the insufficient enrollment due to the low birth rate. To make the school continuously operate and to enhance the competitiveness, the universities have to avoid the loss of students. In this study, we propose the student transfer problem meaning the loss of students who switch from one university to another. We use machine learning techniques to address the problem. We study different machine learning methods such as decision tree, support vector machine (SVM) and Logistic Regression, and use them to construct a student transfer prediction model. The prediction model can be used to predict whether a student will switch from one university to another. Our results show that the student transfer mode built by the decision tree has the best prediction capability. We also use our prediction model to infer several strategies that university administrative chiefs can use them to cope with the student transfer problem.