DATA MINING UNTUK ANALISA PREDIKSI MAHASISWA NON-AKTIF MENGGUNAKAN METODE DECISION TREE C4.5

UNTARI, DWI (2014) DATA MINING UNTUK ANALISA PREDIKSI MAHASISWA NON-AKTIF MENGGUNAKAN METODE DECISION TREE C4.5. Skripsi, Fakultas Ilmu Komputer.

Preview

PDF - Published Version Download (4013b) | Preview

Abstract

The success and failure of a student's study is a reflection quality college. Non-active student is a student who does not register at the beginning of the semester or do not attend a course at least one semester. The existence of non-active students is certainly an effect on the students graduate on time because of the increasing number of non-active students, will increase the number of students do not graduate on time, in addition to the non-active students can potentially increase the number of students drop out. Graduating on time is an element of college accreditation assessment. Therefore, to solve the problem of data mining applied with C4.5 method to search for characteristics potentially non-active students. The design of this study using the CRISP-DM and research using data S1 students of the Faculty of Computer Science of Dian Nuswantoro University. The validation process used is spilled validation, and to testing of model using the confusion matrix. The results showed the best accuracy is 97.60% with 90% of data training. Based on experiments conducted, students with IPS Semester 3 is <2.60, SKS Semester 3 is <20 credits, SKS Semester 4 is empty and IPS Semester 4 is empty greater potential to become non-active students.

Actions (login required)