Predictive analysis of the supply chain management using Machine learning approaches: Review and Taxonomy

Abstract

Currently, there are many literature reviews on the application of predictive analytics in Supply Chain Management (SCM). However, most of them focus only on some specific functions in supply chain management, including Procurement, Demand Management, Logistics and Transportation, or purely technical aspects. The purpose of this paper is twofold: first, it aims to provide an overview of the outstanding supply chain management functions (SCMF) that apply predictive analytics; and second, to highlight practical approaches, algorithms, or models in SCM via a comparative review of machine learning approach for aspect-based predictive analysis. For these purposes, details of relevant literature were gathered and reviewed. Accordingly, this article will present the data, algorithms, and models applied in predictive analytics along with its performance, SCM result taxonomy, which includes all the necessary components in the effective implementing of SCMF. Via the result of the recent related publications and papers (2018- 2020), Demand management and Procurement are the two main areas of SCM, in which predictive analytics is often applied. Particularly, accurate demand forecasting and sensing (Demand management) and sourcing risk management and supplier selection (Procurement) are among the foremost applications of BDA-enabled predictive models. This taxonomy not only helps scientists to have a steppingstone to provide more valuable articles in the future but also allows manufacturers to gain an in-depth understanding of these elaborate scenarios and better manage the supply chain management functions (SCMF) via the application of predictive analytics.