A new approach to considering intermediate weights of NDEA

This study addresses the evaluation of systems structured as Decision-Making Units (DMUs) using Network Data Envelopment Analysis (NDEA). Traditionally, NDEA has treated intermediate factors uniformly across stages, which does not accurately reflect real-world scenarios. This research aims to assess the efficiency of a two-stage series system with unequal weights for intermediate products. The model represents a complex combinatorial problem in Non-Linear Programming (NLP), classified as NP-hard. To solve this, a hybrid approach combining Genetic Algorithms (GA) with NDEA is introduced. This flexibility in handling intermediate weights enhances the model's realism, aligning it more closely with actual DEA applications. The proposed methodology is tested on non-life insurance companies in Taiwan, and the results are compared against existing frameworks, showcasing its practical utility.

کلیدواژه ها:

Network data envelopment analysis ، Two-stage system ، Intermediate weights ، Genetic algorithm ، GA-NDEA algorithm