Treffer: An advanced algorithm-based manufacturing quality forecasting model.

An exponential rise in data availability from manufacturing stages has been caused by the introduction of smart factories and industry. Utilizing the data through supervised machine learning (ML) methods, particularly classification offers a promising avenue for predictions that can significantly mitigate uncertainty, leading to substantial cost savings by preemptively addressing potential issues related to recalls, packaging, and transportation. The application of classification techniques to anticipate overall production batch compliance has received very little attention, despite a wealth of research on quality prediction in particular manufacturing processes. Paper introduced a novel approach integrating Adaptive long short-term memory network-based Vulture-Optimization (ALSTM-VO) algorithm to enhance manufacturing quality forecasting. The suggested model shows remarkable accuracy of 94.5%, recall of 94.3%, and precision of 97.6% in forecasting batch compliance quality through an analysis of a production line for many models of appliances. The results highlight how important feature engineering and dataset understanding are for creating efficient classification models for quality control. [ABSTRACT FROM AUTHOR]