Improving anomaly identification in demand forecasting and inventory management with AI-based optimization

Abstract

Numerous intelligent applications have been made possible in many different areas because to the development of technologies like data mining and machine learning. Supply chain management and communication are two important topics of study related to the Global internet. Among these, inventory management (IM) is gaining significance as a crucial component of the complete life cycle management system of supply chains. However, the lengthy supply chain life cycles, the complexity of supply chain management and the ever-changing customer expectations all contribute to the high logistics and communication costs encountered. By improving the IM process, this effort attempts to reduce costs throughout the supply chain's life cycle. Initially, described as a mathematical model, the main goal of the IM process is to maximize revenues while minimizing logistic costs. The Optimized Hierarchical Clustered Sequential Recurrent Neural Network (OHC-SRNN) is presented as the IM technique to solve this problem in order to achieve this aim. When compared to other cutting-edge techniques, the IM model achieves an excellent average accuracy of approximately 90.2% in predicting inventory requirements. This degree of precision has the potential to improve supply chain efficiency by spotting unexpected inventory activities and reducing inventory expenses by approximately 30%.