Applications of Machine Learning in Supply Chain Management: A Forecasting and Optimization Perspective

Abstract

This paper examines how machine learning can be used in supply chain management particularly demand forecasting and inventory optimization. Through the creation of an intelligent supply chain management structure, a Long Short-Term Memory (LSTM)- based prediction model of demand forecasting, and a genetic algorithm-optimized inventory management system are proposed. The experimental evidence indicates that the suggested LSTM forecasting model still higher the accuracy of the prediction by 23.456 percent and the genetic algorithm-based inventory strategy proves to be beneficial as it decreads the overall cost by 18.923 percent relative to the conventional time series techniques. This study gives a theoretical foundation and practical orientation of transformation of supply chain management to digital.