This study investigates the efficiency of serial production lines using advanced tree-based machine learning algorithms, including Random Forest (RF), XGBoost, LightGBM, and CatBoost. By analyzing various performance indicators such as throughput, machine reliability, and processing time, we construct predictive models to evaluate production line behavior. The comparative results demonstrate that LightGBM outperforms the other models in prediction accuracy and computational speed. These findings suggest that machine learning techniques, particularly ensemble tree models, can significantly enhance decision-making and operational performance in manufacturing systems.

العنوان

Reports in Mechanical Engineering

الناشر

PKP

بلد الناشر

الولايات المتحدة الأمريكية

Indexing

Scopus

معامل التأثير

None

نوع المنشور

Both (Printed and Online)

المجلد

6

السنة

2025

الصفحات

--