|Based on BAN-PT guide in 2008, on-time graduation rate is one of the instruments of accreditation a college assessment. one of the factors that affect the level of accuracy is a graduate student status is inactive students. Case studies on STMIK AMIKOM Purwokerto showed that the accuracy reached 78.80% passed. Steps can be taken to increase the percentage of students passing the right time is to gather information from the student database. Excavation of information can be done with data mining classification. The purpose of this study was to compare several data mining classification algorithms, namely Decision Tree (DT), Naive Bayes (NB), Artificial Neural Network (ANN), Support Vector Machine (SVM) and Logistic Regression (LR) with a basic evaluation of cross validation and T-Test. The method used is the method of comparison. Based on a comparison of performance and test T, SVM algorithm is the most appropriate algorithm used for prediction in a timely graduation. DT algorithm prediction accuracy rate (98.02%) with the excellent classification category. In addition, based on the results of T-Test is very dominant DT algorithm compared with other algorithms.|
Keyword : algorithms; classification; data mining; student; graduate; right.