Analisis Waktu Kelulusan Mahasiswa Menggunakan Bagging Cart pada Fakultas Sains dan Teknologi UIN Sumatera Utara

Abstract

The percentage of students who graduate on time is one of the benchmarks for the accreditation assessment of a college. Therefore, the timeliness of graduating students is one of the things that must be observed by universities or study programs. This study aims to find out the classification of variables that affect the graduation time of undergraduate students of the Faculty of Science and Technology UIN North Sumatra Medan by using the CART method and to find out the increasing percentage of CART prediction accuracy with the Bagging method. The data used in this study is student data from the Faculty of Science and Technology, North Sumatra State Islamic University medan class of 2015 and 2016. The length of thesis preparation is the most influential variable for graduation time. Classification values are seen from the values of accuracy, sensitivity, and specificity. The CART method provides an Accuracy value of 80.64%, Sensitivity of 60%, and Specificity of 84.61%. While the Bagging CART method obtained Accuracy 87.1%, Sensitivity 66.7%, and Specificity 89.2%. Classification results with Bagging CART can increase classification accuracy by 6.5%. Based on the comparison of the accuracy values obtained, it can be concluded that the application of the Bagging method results in better classification accuracy compared to cart classification trees without Bagging.