<![CDATA[The role of data mining in solving a large data problem includes estimation, prediction, classification, clustering and association. One of the roles of data mining used in this study is the classification to predict the graduation of Bina Sarana Informatika students in 2017 with the provision of passing on time and passing late. The algorithm used is the Decision Tree, K-Nerest Neighbor, Naïve Bayes, Random tree and Random Forest. The number of students graduating in the 1st period of 2017 was 5870 where there were 4998 students graduating on time and 872 graduating late. Performance testing of several algorithms using validation and different tests T-Test The k-NN algorithm has a high accuracy value of 99.98% which is the highest value and the best AUC is obtained by the Naïve Bayes algorithm with a value of 1,000. While the difference test is t-test significance <0.05, then H0 is rejected and if> 0.05 then H0 is accepted, only Decision Three against Naïve Bayes is accepted, otherwise it is rejected.]]>
Copyrights © 2019
