Abstract

Understanding the reasoning behind variation in student academic performance in tertiary institutions has been the concern of many researchers for decades. Numerous studies have used traditional statistical methods to identify factors that affect and predict student performance. Machine learning has been successfully applied to so many domains, thus recently, researchers are employing this paradigm for modeling student academic performance and other related problems in higher education. This work focus at addressing the following: proposing optimal algorithm suitable for predicting students academic performance; designing a framework of intelligent recommender system that can predict students’ performance as well as recommend necessary actions to be taken to aid the students and identifying background factors that affect students’ academic performance in tertiary institution at the end of first year. This research used ten classification models and a multilayer perceptron -an artificial neural network function- generated using Waikato Environment for Knowledge Analysis (WEKA). Each model was built in two different ways: the first was built using the 10-fold cross validation, and the second using holdout method (66% of the data was used as training and the remaining as test). Purposive and selective sampling techniques were used in selecting one thousand five hundred (1,500) enrolment records of students admitted into computer science programme Babcock University between 2001and 2010. Results of the classifiers were compared using accuracy level, confusion matrices and speed of model building benchmarks. The random tree identified as optimal in this work is incorporated into designing a framework of intelligent recommender system. The work shows that identifying the relevant student background factors can be incorporated to design a framework that can serve as valuable tool in predicting student performance as well as recommend the necessary intervention strategies to adopt.