Mining Sebha University Student Enrolment Data Using Descriptive and Predictive Approach

One of the main concerns of higher educational system is evaluating and enhancing the educational organization. For achieving this quality objective the organizations need deep knowledge assess, evaluate and plan towards better
decision making process. Data mining techniques are analysis tools that can be used to extract meaningful knowledge from large databases. This study presents
applying data mining in the field of higher educational especially for Sebha University in Libya. The main contribution of the study is an analysis model that can be used as a decision support tool. It acts as a guideline or roadmap to identify which part of the processes can be enhanced through data mining technology and how the technology could improve the conventional processes
by getting advantages of it. Two main approaches were used in this study. Firstly the descriptive statistics, particularly cross tabulation analysis was carried out and presents a lot of useful information within data. Cluster analysis was performed to group the data into clusters based on its similarities. The clusters were also used as targets for prediction experiment. For predictive analysis, three techniques have been used Neural Network, Logistic regression and the Decision Tree. The study shows that Neural Network obtains the highest results accuracy among the three techniques.