Perancangan Sistem Klasifikasi Mahasiswa untuk Prediksi Performa Mahasiswa Menggunakan Naïve Bayes Classifier

Abstract

Abstract – One of the important aspects in determining student performance in the lecture process is the timeliness in completing their studies. The study period of each student can be caused by various factors. By using the Naïve Bayes Classifier algorithm, it is possible to predict student performance in the Information Systems Study Program, UIN alauddin Makassar. The attributes used are gender, 1st semester GPA, 2nd semester GPA, entry path, high school major, school type, economic status, working status, marital status and organization. The student sample, namely alumni data from the 2011 to 2017 batches was used as training and testing data totaling 275 data, while student data from the 2019 and 2021 batches as target data were 153 data. The testing process will be carried out using 10 Fold Cross Validation method, and Confusion Matrix method. The results of the test show that the average performance of the Naïve Bayes model has an accuracy of 70%, precision of 69%, recall of 70% and f1-score of 67%. In this research, the system design uses Agile Method which is optimized with Laravel Framework. The Laravel framework is a framework that has a concise and efficient MVC (Model, View, Controller) concept used in developing a web application and has many helpers that can facilitate and assist in the application creation process.   System testing using Black Box testing techniques.

Keywords: Predict, Naïve Bayes, student, 10 fold cross validation, confusion matrix.