J-SAKTI (Jurnal Sains Komputer dan Informatika)
Vol 8, No 1 (2024): EDISI MARET

Artificial Neural Network Untuk Prediksi Kelulusan Calon Peserta Didik Baru (Studi Kasus: MAN 1 Padangsidimpuan)

Gaja, Rizqi Nusabbih Hidayatullah (Unknown)
Karseno, Doni (Unknown)
Rijal, Amir Salim Khairul (Unknown)
Juliantho, Dwana Abdi (Unknown)



Article Info

Publish Date
30 Mar 2024

Abstract

There is fierce rivalry amongst schools as a result of competitiveness in the educational field. Therefore, the processing of an educational institution must keep up with technological advancements. Artificial intelligence (AI) and intelligent systems, sometimes called AI systems, are used to simulate human-like critical thinking and intelligent behavior. A popular technique for categorization and prediction is the artificial neural network. MAN 1 Padangsidimpuan faces the challenge of determining whether new students will graduate because each year's quota of applications is exceeded. In this study, potential new students at MAN 1 Padangsidimpuan will have their graduation dates predicted, and the degree of prediction accuracy will be assessed. Artificial Neural Networks are the research approach employed in this study. The steps that are completed include problem formulation and identification, literature review, data gathering, pre-processing, processing, and assessment. Data about potential new students for the academic year 2023–2024 was utilized in this study. The study's findings demonstrate that the neural network model produces outcomes that are 6-6-2 (6 input neurons, 6 hidden layers, and 2 output neurons). 97.31% was the greatest accuracy performance level attained, while 90.30% was the lowest. 615 people made the most accurate predictions, while 569 people made the fewest accurate ones. There were 17 forecasts that were wrong the least and 63 wrong predictions the most.

Copyrights © 2024






Journal Info

Abbrev

jsakti

Publisher

Subject

Computer Science & IT Control & Systems Engineering Decision Sciences, Operations Research & Management Energy

Description

J-SAKTI adalah jurnal yang diterbitkan oleh LPPM STIKOM Tunas Bangsa yang berfokus di bidang Manajemen Informatika. Pengiriman artikel tidak dipungut biaya, kemudian artikel yang diterima akan diterbitkan secara online dan dapat diakses secara gratis. Topik dari J-SAKTI adalah sebagai berikut (namun ...