A Distance-Based Multi-Model Fusion Method for Leaf Group Formula Raw Material Substitution Recommendation in Yunnan Cigarette Products

This research aims to meet the practical needs of leaf group formula maintenance in cigarette production by utilizing machine learning to enhance the scientific and timely use of raw materials. A novel raw material substitution recommendation algorithm based on distance multi-model fusion is proposed. The algorithm comprises five spatial distance algorithm units and a voting decision mechanism, which together form a complete recommendation index mapping method. The results demonstrate that this method provides stable recommendation capabilities and offers a novel, efficient approach for formula maintenance personnel.