Exploring applications of Machine Learning for supply chain management

In the current economic context, due to globalization, the trend of outsourcing/off shoring and the flexibility of distribution processes, companies have moved towards sourcing products from distant markets around the world in order to save production costs. This, in turn, has resulted in longer lead times for the production and supply of products, which has increased the uncertainty of demand. In fact, all supply chain processes are based on demand forecasting data, so methods to improve forecasting can increase the efficiency of the supply chain. Machine learning techniques, have recently emerged as an alternative to traditional methods, and have proven to be capable of solving demand forecasting problems. In this paper, we explore the bibliography of recent research of Machine Learning applications to supply chain and in particular demand forecasting. A total of 79 articles were reviewed and collected of two databases, ScienceDirect, and Scopus. The study demonstrated the power of machine learning algorithms over traditional demand forecasting model, and showed that neural networks where one of the most used algorithms in demand forecasting. Finally, the strength of machine learning techniques in solving demand forecasting problems in different areas, asserts its remarkable influence in improving the efficiency of the supply chain. This can encourage both decision makers and stakeholders to plan corrective actions based on machine learning applications to supply chain and demand forecasting.