PENERAPAN JARINGAN SYARAF TIRUAN UNTUK MENGUKUR TINGKAT KORELASI ANTARA NEM DENGAN IPK KELULUSAN MAHASISWA

Abstract

Currently, the Informatics Engineering undergraduate program of University of X (PSTI-X) uses the SLTA's EBTANAS Score (NEM) for admission selection criteria, especially for those coming who apply through fellowships. Due to the fact that the high score of NEM does not guarantee a graduate with high academic achievement, thus this study is intended to test the correlation between the composition of the high school's NEM with the grade point average of PSTI-X graduate using artificial neural network with feed-forward back propagation approach (JST-PB). It is expected to be used as a guideline for determining the composition of the SLTA's NEM score in the selection of new admissions, particularly via the fellowship. The results showed that JST-PB only has 64% success rate in the mapping of NEM composition and academic achievement of graduates. These results reinforce the results of previous studies using correlation analysis and multiple linear regressions that NEM could not be used as an absolute reference in the new admission selection system. The result of this study provides recommendations that PSTI-X needs to review new admissions system, particularly those through the fellowship.