Data Mining Approach for Educational Decision Support

Sinta Septi Pangastuti (Statistic Department, Universitas Padjajaran)
Kartika Fithriasari (Statistic Department, Institut Teknologi Sepuluh Nopember)
Nur Iriawan (Statistic Department, Institut Teknologi Sepuluh Nopember)
Wahyuni Suryaningtyas (Mathematics Education, Muhammadiyah University of Surabaya)

EKSAKTA: Journal of Sciences and Data Analysis
VOLUME 2, ISSUE 1, February 2021

data mining techniques in education sector have begun to evolve, along with the development of technology and the amount of data that can be stored in an education database storage system. One of them is a database of Bidikmisi scholarships in Indonesia. The Bidikmisi data used in this study will be classified using classification data mining technique. The technique that used in this study is random forest in combination with boosting algorithm and bagging algorithms. These algorithms also combine with SMOTE algorithm to handling the imbalance class in dataset. Based on the performance criteria G-mean and AUC, the algorithm combines with SMOTE tended to be better. The classification accuracy of each method being more than 90%