Predicting and managing risk interactions and systemic risks in infrastructure projects using machine learning

Infrastructure projects often encounter performance challenges, such as cost overruns and safety issues, due to complex risk interactions and systemic risks. Existing literature treats risk interactions and systemic risks separately and relies on models that struggle with nonlinearities, adaptability, and practical applications, leading to suboptimal risk management. To address this gap, this paper uses machine learning (ML) algorithms to analyze historical project data and predict the impacts of risk interactions and systemic risks on future projects. The results show that ML-based models provide accurate and practical data-driven predictions of project performance under risk interactions and systemic risks. These findings are valuable for infrastructure project managers seeking to improve risk mitigation strategies and project outcomes. The paper lays also the foundation for future research on leveraging advanced predictive analytics in managing complex project risks more effectively. • Risk interactions and systemic risks are imperative for efficient risk management. • Existing models are impractical and involve complex simulations. • Data-driven model for risk interactions and systemic risk assessment is introduced. • Practical risk analysis approach to address overlooked risk assessment dimensions. • Intuitive model architecture for risk prediction.