Stacked Differential Evolution for Data Envelopment Analysis

Data envelopment analysis (DEA) is a popular method for the complex evaluation of the efficiency of entities that perform similar operations with different ratios of inputs and outputs. In this work, we use a modern version of the differential evolution algorithm, j2020, to analyze DEA instances. It is used in the context of a novel DEA solution strategy in which the high-level task of DEA instance analysis is transformed from a set of low-dimensional optimization problems into a lower number of higher-dimensional ones. It allows the evaluation of the efficiency of multiple DMUs in a single run, decreases the number of individual optimization problems that need to be solved to assess complete DEA instances, and as a result provides an opportunity for the use of optimization algorithms that can efficiently operate in high-dimensional search spaces.