Penerapan Jaringan Saraf Tiruan dengan Metode Pembelajaran Backpropagation untuk Mengetahui Tingkat Kualifikasi Calon Siswa pada Sistem Informasi Penerimaan Siswa Baru di MAN 2 Banjarnegara

Abstract

The selection process of the new students at Madrasah Aliyah Negeri 2 Banjarnegara aims to get the best prospective students to improve the quality and quantity of the achievement of Madrasah Aliyah Negeri 2 Banjarnegara they needs to know the qualifications of the applicant. To help the school to know the qualifications of prospective students required an application that is able to analyze the ability of prospective students, that will assist the school in predicting the qualifications of prospective students. The application is designed using artificial neural network processing with backpropagation learning algorithm. From the experiments performed 6 times used parameter : target error 0,001, the maximum epoch 10000 and learning rate from 0,3 to 0,8 obtained satisfactory results with 64 kinds of patterns tested system can recognize 100% patterns with MSE smaller than 0,001 and when they tested with 100 students sample data that they got from the documents the school system can recognize 100% of the data. Based on the experimental magnitude of learning rate affects the number of iterations to get the MSE, the greater the learning rate value the smaller value of iterations required to obtain a smaller MSE than the target error