Supporting Supply Chain Risk Management: An Innovative Approach Using Graph Theory and Forecasting Algorithms
Purpose: The article aims to develop a tool supporting risk management in the supply chain. Design/Methodology/Approach: Graph theories were employed for data analysis. The study introduces a method for forecasting the behavior of sales risk values over a specific time horizon using a time series approach. Several metrics were utilized, starting with degree centrality, which assumes that crucial graph nodes have numerous connections. Betweenness centrality was also considered, assuming that key nodes link other nodes and measuring how often a vertex lies on paths between other vertices. Additionally, the PageRank algorithm, developed by Google, was applied. Findings: The study produced an analysis of supply chain demand forecasting using the Monte Carlo method. Preparing such a forecast allows you to check many possible outcomes of the decision-making process. It can be used to assess the impact of risk, which in turn allows for better decision-making in conditions of uncertainty. Practical Implications: The proposed method for forecasting sales values over a particular time horizon enables consideration of demand prediction within the supply chain, including associated risks. Originality/Value: What is new is the use of graph theory to review supply chain risks and algorithms for applications supporting supply chain management.