IDENTIFIKASI LAMA STUDI BERDASARKAN KARAKTERISTIK MAHASISWA MENGGUNAKAN ALGORITMA C4.5 (Studi Kasus Lulusan Fakultas Sains dan Matematika Universitas Diponegoro Tahun 2013/2014)

Abstract

Based on academics regulation No. 209/PER/UN7/2012, the study period of student in Diponegoro University has been scheduled that is 4 years. In this thesis the graduation status of students that graduate under or equal to 4 years categorized as graduate on time, meanwhile students that graduate over 4 years categorized as graduate out of time. Thus background, students identification to classfy students that graduate on time can be known using Algorithm C4.5 based on gender, majors, GPA, organizational experience, part time experience, scholarship, students origin and pathways scholar. Algorithm C4.5 contructs decision tree that can handle with missing values, continues attribute and pruning trees. Results of the Algorithm C4.5 that constructs a decision tree shows that majors is selected as a root node, the accuration of the Algorithm C4.5 is 84,475% and the number of the nodes are 20 nodes where 13 nodes are leaf nodes. The students profile that identified graduate on time are students of Physics who had received scholarship and a woman; students of Chemistry with GPA > 3,06; students of Statistics with GPA > 3,43 from SNMPTN also PSSB and students of Mathematics with GPA > 2,96. Keywords: Study Period, Algorithm C4.5, Decision Tree.