PENINGKATAN AKURASI PREDIKSI PENJURUSAN SISWA SMK DENGAN OPTIMASI JARINGAN SYARAF TIRUAN BACKPROPAGATION

Abstract

The development of artificial intelligence (AI) technology has had an increasingly significant impact on various industries, including education, particularly in terms of data processing and decision making. However, in reality, students' choice of major is often determined without proper and measurable analysis, which means that students' potential is not always in line with their chosen major. The mismatch between academic abilities and chosen fields of study is one of the problems arising from this situation. To address this issue, this study predicts majors based on subject grade data using Artificial Neural Network techniques and the Backpropagation algorithm. Backpropagation was chosen because it can produce more accurate predictions by gradually learning data patterns through a directed learning process. This approach significantly improves prediction accuracy based on model training and testing results, making it a useful tool for more objective, flexible, efficient, adaptive, and data-driven decision-making in optimally selecting majors for students to support their overall and sustainable academic success.