Predicted Student Study Period with C4.5 Data Mining Algorithm

Abstract

Data of alumni from 2012 to 2015 found that the average percentage of students graduating on time was 22%. The comparison between the number of students who graduate on time and new students who enter each year is not comparable, therefore a study is needed to find out the factors that affect student graduation and to prediction of the graduation period of the student through data mining research using the C4.5 algorithm. The data tested was student alumni data from 2012 to 2015. The instruments studied include study period, academic year, GPA, corner focus, gender, intensity of work during college, type of thesis, intensity of campus internal organization, intensity of external organization of campus, UKT group, scholarship status, pre-college education, hobby intensity, intensity of game play, academic competition participation status, non-academic competition participation status, and availability of facilities and infrastructure. The best test results using percentage-split 75% obtained 83.33% accuracy as well as the rules contained in the decision tree.

Keywords

• c4.5 algorithm
• data mining
• predicted study period