KLASIFIKASI PENERIMAAN PESERTA DIDIK BARU MENGGUNAKAN ALGORITMA NAÏVE BAYES DENGAN SMOTE PADA SMKN 1 JAMBLANG

Abstract

Abstract: Each educational institution has its own way in peroses acceptance of new learners. In order for the acceptance of new didi participants effectively and efficiently, it is necessary to analyze the data of new learners' admissions. Data mining techniques are one way that can be done to classify new learners' data. Class imbalance problems usually occur when classifying, where a classifier tends to classify the majority class and ignores the minority class. To overcome this problem can be used two approaches, namely, the sample approach and the algorithm. in overcoming the problem of unbalanced data on the admission data of new learners SMKN1 Jamblang using the sampling technique approach. Sampling techniques that are commonly used in overcoming class imbalance problems are over-sampling, under-sampling, and a combination of both. The completion of class imbalance is seen based on accuracy, sensitivity, and specificity with naive bayes classification method combined with SMOTE algorithm in the new student admission dataset in SMKN 1 Jamblang. The results of the study conducted at SMKN1 jamblang on 638 new student admission data in SMKN 1 Jamblang were classified into two classes, namely 274 accepted classes and 265 classes were not received with an accuracy of 92.58%. The results of the completion of class imbalance against 638