Comparison of Adaboost Application to C4.5 and C5.0 Algorithms in Student Graduation Classification

Abstract

Students become a benchmark used to assess quality and evaluate college learning plans. Therefore, students who graduate not on time can have an effect on accreditation assessment. The characteristics of students who graduate on time or not on time in determining student graduation can be analyzed using classification techniques in data mining, namely the C4.5 and C5.0 algorithms. The purpose of this study is to compare the application of the Adaboost Algorithm to the C4.5 and C5.0 Algorithms in the classification of student graduation. The data used is the graduation data of students of the Statistics Study Program at Tanjungpura University Period I of the 2017/2018 Academic Year to Period II of the 2022/2023 Academic Year. The analysis begins by calculating the entropy, gain and gain ratio values. After that, each data was given the same initial weight and iterated 100 times. Based on the classification results using the C5.0 Algorithm, the attribute that has the highest gain ratio value is school accreditation, meaning that the school accreditation attribute has the most influence in the classification of student graduation. The application of the Adaboost Algorithm to the C5.0 Algorithm is better than the C4.5 Algorithm in classifying the graduation of students of the Untan Statistics Study Program. The Adaboost algorithm was able to increase the accuracy of the C5.0 Algorithm by 12.14%. While in the C4.5 Algorithm, the Adaboost Algorithm increases accuracy by 10.71%.