Enhancing Supply Chain Performance through AI: A Predictive Analytics Approach Using Weighted Regularized Extreme Learning Machine Model

Abstract

Abstract: Supply chain management (SCM) is a crucial component of any competitive strategy aimed at increasing organizational profitability and productivity. The discipline of SCM has a wealth of literature on strategies and technologies for effective SCM. There has been a deluge of academic and professional activity in recent years devoted to metrics and evaluations of organizational effectiveness. Training the model, selecting features, and preprocessing are its main components.  There are three types of normalization used in data preprocessing: min-max, z-score, and decimal scaling. The most accurate method is z-score normalization. To pick features, we employ the sine-cosine algorithm. For this purpose, we trained the model using the WRELM framework.  It makes ELM and RELM look antiquated in comparison. According to the numbers, the accuracy rate is 96.20%.