A hybrid genetic algorithm and DEA approach for multi-criteria fixed cost allocation

This paper proposes a hybrid genetic algorithm and data envelopment analysis framework for solving the fixed cost allocation (FCA) problem. The proposed framework allows managers to incorporate different FCA sub-objectives for efficient and inefficient decision-making units (DMUs) and solves the FCA problem so that the total entropy of resource allocation for efficient DMUs is maximized, and correlation between resource allocation and efficiency scores of inefficient DMUs is minimized. The FCA sub-objectives and solutions are kept consistent with the overall management objective of rewarding efficient DMUs by allocating to them fewer fixed cost resources. We illustrate the application of our approach using an example from the literature. The results of our study indicate that the solution values obtained in our study are superior to those obtained in other studies under various criteria. Additionally, the relative gap between the solution obtained using our procedure, and the upper bound on the optimal value is approximately 1%, which indicates that our solution is very close to the optimal solution.