A Robust Framework for Generalized Leader-Follower Network Data Envelopment Analysis Under Uncertain Data

This research focuses on assessing the efficiency of network decision-making units (DMUs) in uncertain environments using network data envelopment analysis (NDEA). Traditional NDEA models are effective in analyzing multi-stage systems but face challenges in uncertain conditions due to their dependence on precise data. To address this issue, the study introduces a leader-follower NDEA model with various returns to scale assumptions, enhancing its flexibility. A robust optimization approach utilizing convex uncertainty sets is incorporated to manage data uncertainty, ensuring reliable performance evaluations even in dynamic and imprecise scenarios. This methodology maintains the accuracy of efficiency analysis despite ambiguous or incomplete data. The proposed robust leader-follower NDEA model is validated through a real-world case study involving investment companies in the Iranian capital market. The results highlight the framework's resilience and capability to handle data ambiguity, offering valuable insights into system efficiency and identifying performance bottlenecks. This research presents a practical and dependable tool for performance evaluation in complex and uncertain operational environments.