Predictive Machine Learning Modeling for ERP Global Order in Supply Chains Based on Hybrid Attention SNN Approach

Abstract

Management of supply chains is of the utmost significance to today's successful businesses operating in the global market. Innovation in supply chain solutions, including methods to make them quicker, better, and cheaper, is a constant focus for companies as they strive to maintain a competitive edge. Improving consumer happiness is essential, and one way to do this is by reliably delivering on promises. The ability of a corporation to fulfill its promises in this particular scenario is contingent upon its ability to plan and execute effectively. The three main components are feature selection, model training, and preprocessing.  The process of preprocessing involves transforming the format of unstructured data in order to make it more understandable. Data mining also relies on this step, since raw data is useless without it.  Among the two methods used in feature selection PCA and DPCA—DPCA produces superior results. We used the Attention-SNN framework to train the model for greater precision.  With an accuracy of approximately 96.48%, the suggested method surpasses rival approaches, such as SNN and Attention mechanism.