PREDIKSI MASA STUDI SARJANA MENGGUNAKAN DATA MINING PADA PRODI S1 MATEMATIKA FMIPA UNIVERSITAS RIAU

Abstract

Prediction of undergraduate study period is influential in helping the study program to monitor the development of studies from its students and prevent the existence of students that will result in the perfomance of the study program decreases. It can also help the study program in producing a skinned scholar who will be useful during the reaccreditation process. This research aims to find out the predictions of the length of study period of students using the web-based Naive Bayes method. The prediction of the undergraduate study period consists of two stages, namely the stages in data processing using the Naive Bayes method and the stages in system creation. The stages of data processing using the Naive Bayes method consists of data collection, data cleansing, data integration, data selection, data transformation and Naive Bayes stages. While the stages in the creation of the system consist of requirment, design, implementation and testing. Users can input variabels that are used to make prediction, then the system will provide the results of student study period predictions based on calculations of Naive Bayes method. The variabels used are gender, GPA semester IV, parental work and parental income also the resulting predictions are study period on time or not. The system is designed using UML diagrams with PHP programming languages and MySQL databases. The results of system testing conducted using confusion matrix againts 28 undergraduate study period data with on time and not on time labels resulted in an accuracy rate of 78.57%.