Analyzing Academic Achievement of CAS's Students Using Data Mining

Massive information can be collected from students' data in order to produce knowledge. The educational institutions collect students' data such as academic information, demographic, and personal traits. The data collected
based on these variables used to predict the students' academic achievement. On this study, the respondents are students who have graduated within the period of six months in the year 2006, 2007 and 2008. Two data mining techniques for analyzing and building the classification model for students' achievement in College of Arts and Sciences (CAS), Universiti Utara Malaysia (UUM) are presented. Initially, the relationship and correlation between students' cumulative grade point average (CGPA) with academic background, demographic, entry qualification, sponsorship and interpersonal skills, students' achievement are analyzed. For model building purposes, final CGPA has been used as a target. The analysis conducted using
Multinomial Logistic Regression and Neural Network found that, gender, entry qualification, language qualification (Bahasa Malaysia and English), family income, sponsorship, analytical and analysis skill as well as teamwork are all the best predictors contributed to students' performance.
The result obtained through this study can be used to help the management of CAS to make certain decisions and to predict the outcome of current and future students.