Predictive Modeling for Student Performance Data Using Decision Tree and Support Vector Machine

• Main Authors: Maryam Khanian, Najafabdi, Sarasvathi, Nagalingham, Sayed Mojtaba, Tabibian
• Format: Article
• Language: English
• Published: INTI International University 2019
• Subjects: L Education (General); LB2300 Higher Education; QA75 Electronic computers. Computer science
• Online Access: http://eprints.intimal.edu.my/1329/1/ij2019_08.pdf; http://eprints.intimal.edu.my/1329/; http://intijournal.newinti.edu.my

This paper is aimed to present a conceptual understanding that summarizes higher education analytics lifecycle. This paper explores the establishment of new architecture of technologies, experts, standards, and practices in the complex data infrastructure projects among higher education institutes. The research on higher education analytics converges with the demands from industry to improve the learning education systems by considering the teaching and learning analytics capabilities enhancing the efficiency of higher education. The exploitation of massive volume campus and learning information could be a crucial challenge for the planning of campus resources, personalized curricula and learning experiences. In the field of higher education, institutions look to a future of the unknown and vast speed advancement of technology. Moreover, with more strategic data solutions used in decision making with the over increasing social needs and political changes at national and global, competition within and among universities increase. Higher education needs to expand local and global impact, increase financial and operational efficiency, create the new funding models in a changing economic climate and respond to the greater accountability demands to ensure the success of organizational at all levels and stay on top of the ranks. Research on higher education institutes is also important because it enables maximum benefit and perceptions on students’ performance and learning trajectories to be determined as these two are important in adapting and personalizing curriculum and assessment. The findings of this paper provided a view about modeling students’ performance classification by Machine Learning models and to identify which of the predictors in the dataset contribute towards good prediction on the students’ performance.