Akademik Veri Yönetim Sistemi
1st International Conference on Optimization and Data Science in Industrial Engineering, ODSIE 2023, İstanbul, Türkiye, 16 - 17 Kasım 2023, cilt.2204, ss.86-103, (Tam Metin Bildiri)
Today, supplying the right quantity of materials and ensuring their availability are key factors in reducing inventory costs and improving the competitiveness of the industry. Companies need to hold large quantities of inventory to meet future demand. Proper inventory control is necessary to easily monitor and manage inventory levels, meet customer needs by controlling inventory levels, and balance items to be procured. To this end, a multi-criteria inventory classification (MCIC) with the integration of an analytic hierarchy process (AHP) and machine learning algorithms is proposed to examine the inventories of a manufacturing industry and create an applicable and effective inventory management system. First, the classes of inventory items are identified by ABC analysis using the AHP method. Then, their performance in inventory classification was evaluated using machine learning (ML) algorithms such as Decision Tree, Naive Bayes, Random Forest, and Support Vector Machine. The results show that the random forest algorithm is more effective than other methods in classifying inventory items.