PREDIKSI PENENTUAN BAKAT DAN MINAT SISWA DENGAN MENGGUNAKAN METODE CART (CLASSIFICATION AND REGRESSION TREE) (STUDI KASUS: Madrasah Aliyah Al Hadi Girikusuma)

Abstract

Abstract

The determination of talent and interest is the determination of majors that are introduced by Senior Hight School. Madrasah Aliyah Al Hadi Girikusuma is a private madrasah of Senior Hight School and vocational high school that also implements the process of determining talent and interest for their students. Participants in the determination of talents and interests followed by the participants of class X with the number of participants as many as 124 students. The number of students and the process that is still manual makes the old and less efficient in the determination of talents and interests. This quantitative research predicts the determination of student's talents and interests by using CART (Classification And Regression Tree) method and using Repied Miner 5.3.013 tools. which will define some labels, resulting in 3 status predictions, namely IPA, IPS and Religion. Processed data using 3 variable indicators include knowledge variables, practice variables and attitude variables from each subjects under the control. The result of CART classification with accuracy from experiment using Fold Cross validation method with sampling type using shuffled sampling yield accuracy of 73.40%, precision of 82.43% and recall of 80.26%.