Prediksi Penentuan Program Studi Berdasarkan Nilai Siswa dengan Algoritma Backpropagation

Abstract

In continuing higher education, the selection of study programs for students is considered difficult and confusing in choosing the right study program. Data Mining technique is a process of finding new knowledge from a set of databases that can help predict the selection of the appropriate study program. This prediction uses the Backpropagation Neural Network method which aims to assist in the selection of subjects according to student scores with a sample of 50 data to be trained and tested. There are seven input variables used, two hidden layers with varying number of nodes, and outputs that will be used as references in the selection of courses. In this study, the Artificial Neural Network Backpropagation method uses classification performance with Rapidminer software on the target that produces the greatest accuracy is 77.42%.