Application of C5.0 Algorithm in Prediction of Learning Outcomes in Calculus Subject

Abstract

Calculus is one of the basic subject that must be studied at the computer science faculty of the informatics engineering study program. For some students, especially in the Faculty of Informatics Engineering, calculus is a subject that is considered quite difficult, even though this subject is important for them. And the resulted for some students having to repeat this subject. For this reason, predictions of calculus learning outcomes are carried out by applying the data mining process and using the C5.0 method for the prediction process based on the classification concept that will be carried out. This study applies the Cross Industry Standard Process for Data Mining (CRISP – DM) methodology with the C5.0 algorithm. The results are in the form of a decision tree (Decision tree) and the rules in it using the attributes of guardian, number of family members, status of residence, internet, activity, desire to continue study, the last education of parents (father and mother), parents' occupations, grades on assignments, UAS, and UTS. The C5.0 algorithm is able to predict the results of learning calculus. The evaluation results show that the applied C5.0 algorithm has an accuracy of 95%.

Keywords

• data mining
• prediction
• C5.0 algorithm
• CRISP-DM
• student achievements