An Intelligent Multi-Criteria Network Algorithm for Modeling Interdependencies among Risk, Competitiveness, and Sustainability in Industrial Systems

Industrial systems increasingly operate under multidimensional uncertainty, where risk exposure, competitiveness capability, and sustainability performance interact through complex interdependencies. Conventional multi-criteria evaluation models often assume hierarchical or independent relationships, limiting their ability to capture reciprocal influence structures. This study proposes an Intelligent Multi-Criteria Network Algorithm (IMNA) to model these interactions using a directed weighted network framework. The model incorporates 51 criteria grouped into risk (20), competitiveness (22), and sustainability (9) clusters. After normalization and adaptive threshold filtering, weak connections are removed to improve structural clarity. Spectral centrality analysis identifies 13 dominant variables that drive system interactions, particularly those related to supply chain stability, market conditions, and environmental management. Stability testing under perturbation levels of 5%, 10%, and 20% yields rank shifts of 0.52, 0.74, and 1.18 (overall mean 0.81); the rank shift at 10% noise of 0.74 falls well below one position, indicating robust network inference. The proposed framework effectively reduces system complexity while revealing key structural relationships for decision support.