<![CDATA[Informatics is one of the study programs at STIKI Malang which has quite large student data both data on active students and students who have graduated. Every year students who graduate on time are far less than new students, which will affect the quality of students and also influence to the accreditation of STIKI Malang. Therefore, the purpose of this study is to make an application to predict the duration of study and the predicate of student graduation by applying supervised learning techniques. Output predictions for the period of study are on time, late, or not graduated and output for predicate are summa cum laude, cum laude, very satisfying, or satisfying. The application method used in supervised learning for prediction is a Naïve Bayes algorithm. This is used to analyze data, especially in the pattern recognition process, predicting the period of study and the predicate of graduation. Before entering the calculation phase of the Naive Bayes algorithm, the Correlation Feature Selection is used to select relevant features that function to improve the accuracy of the system. The results of the study based on 10 Fold Cross Validation testing indicate that the application can be used to assist informatics study programs in order to find strategic information related to the duration of study and the predicate of student graduation with an accuracy of 77.19% and 87.65%.Keywords: Prediction, Naïve Bayes, Study Periods, Correlation Feature Selection, 10 Fold Cross Validation]]>
Copyrights © 2019
