A Study on Clustering Analysis for Students＇Choices of Department Preference

• Main Authors: Chen, Yin-Shao, 陳胤少
• Other Authors: Yu, Chang-Yung
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/96490196435930901575

碩士 === 靜宜大學 === 財務與計算數學系 === 104 === Along with the birth population decline, the enrollment of universities becomes more and more competitive. To understand the optional voluntary strategy for examinee could be helpful for the strategy of department admissions. Examinees are allowed to choose up to six departments for the second stage examination. Therefore, basket analysis could be applied if we treat the departments examinee choose as goods. Data mining technology has been widely used in various fields, its object is to find out the potential valuable information from a large number of transaction data. The association rules method is the most often used technology in basket analysis to identify the association between trading projects. In addition, cluster analysis is used to establish clusters with high intra-cluster similarity and low inter-cluster similarity. We discuss how to combine these two methods and applied to the data announced by 2016 university admission committee. We analyzed the common competitive cluster structure of departments, and also compared the difference with data from 2014 and 2015. We expect this method can be used to validate the department admissions policy.