Framework for Optimising Supply Chain Analysis Using Machine Learning

An algorithmic framework is presented to predict sales, late deliveries and fraudulent transactions in a supply chain using machine learning (ML) algorithms. The framework utilizes past data to recognize regularities, inclinations, and irregularities within the supply chain, enabling precise forecasting of future occurrences. Several algorithms were used out of which random forest and logistic regression were best performing. Supply chain managers can gain insights from this framework to empower their business and mitigate risks. Supply chain can be optimised and revolutionised by adapting the proposed framework to various supply chain scenarios.