A Bayesian network approach to production decisions by incorporating complex causal factors

Handling uncertainty is a key aspect of production management. This paper employs a Bayesian network (BN) approach to improve uncertainty analysis in production decision-making by identifying common causal factors that may affect one or more components of an integrated supply chain problem. A Bayesian network approach with learning and analytical features is proposed, allowing production managers to make risk-based decisions as well as conduct various scenarios and what-if analyses. Additionally, this approach provides managers with the possibility of updating their knowledge about uncertain factors in the light of new data. The proposed model is applied to a vaccine manufacturing and distributing chain, revealing that the company will fail to gain a profit unless it reduces costs, enhances staff quality, or acquires more customers with higher demand. • An intelligent model to incorporate uncertainties into production decision-making problem. • A Bayesian network approach with learning and focusing on common and complex causal factors. • Examining the validation of decision-making approach through a real-case application. • Analyzing different scenarios to find the root causes affecting performance and the best decision.