Prediction Of Students’ Learning Study Periode By Using Random Forest Method (Case Study: Stikom Bali)

• Main Author: I Made Budi Adnyana
• Format: Article
• Language: Indonesian
• Published: LPPM Universitas Potensi Utama 2016-10-01
• Series: CSRID Journal
• Subjects: prediction; study time; random forest
• Online Access: http://csrid.potensi-utama.ac.id/index.php/CSRID/article/view/95
• ISSN: 2085-1367; 2460-870X

Graduation on time is one of the assessment elements of the college accreditation. Furthermore, graduation on time is an important issue because it indicates an effectiveness of college. Academic division of STIKOM Bali face many difficulties on predicting student graduation time because lack of information and analysis. Predictions of graduation time can help academic division in making appropriate strategy to shorten the study time. Data mining can be applied on this prediction problems using random forest classification methods. Random forest is a collection of of several tree, where each tree dependent on the pixels on each vector that selected randomly and independent .Sample data obtained from academic division of STIKOM Bali. This research use sample data of last 2 years graduated students, such as IPK, SKS, the number of inactive, and study time. The classification output consists of 2 class, “graduate on time” and “graduate over the time”. From the experimental result, 83.54 % accuracy value obtained.