A hybrid machine learning-enhanced MCDM model for transport safety engineering.

Area of Science:

• Engineering
• Computer Science
• Operations Research

Background:

• Multi-criteria decision-making (MCDM) is crucial for transport safety engineering.
• Traditional methods face challenges with data uncertainty and computational efficiency.

Purpose of the Study:

• To propose a hybrid machine learning-enhanced MCDM model for robust decision recommendations.
• To improve the reliability and efficiency of decision-making processes in transport safety.

Main Methods:

• Integration of Distance Correlation-based Criteria Importance through Intercriteria Correlation (DCRITIC) for criteria weighting.
• Application of Weighted Aggregated Sum Product Assessment (WASPAS) for decision aggregation.
• Utilization of a graph-based machine learning technique to enhance K-means clustering for robust centroid selection, creating the DCRITIC-WASPAS-K-means model.

Main Results:

• The DCRITIC-WASPAS-K-means model demonstrated improved robustness and reliability in decision outcomes.
• The machine learning integration reduced uncertainty associated with initial centroid selection in K-means clustering, decreasing iterations and runtime.
• A case study in the Organization of American States (OAS) region validated the model's practical utility and superior performance.

Conclusions:

• The proposed model offers a reliable, scalable, and data-driven tool for strategic planning and resource allocation in transport safety.
• It enhances the credibility of policy interventions by providing consistent decision outputs and effective communication of policy implications.
• Public administrators, policymakers, and government agencies can leverage this framework for improved decision-making in uncertain environments.