THE IMPACT OF SELECTIVE CONVEXITY ON THE STRUCTURES OF PRODUCTION FUNCTIONS IN DATA ENVELOPMENT ANALYSIS

The Data Envelopment Analysis (DEA) technology was proposed at the end of the last century. This technology enables managers to find important characteristics of production units behavior. It also permits to analyze various situations with production units, to compute important characteristics of production units: Marginal rates, efficiency scores, and so on. The Free Disposal Hull (FDH) technology appeared almost at the same time. The production possibility set (PPS) of the DEA models is convex, and the PPS of the FDH models is non-convex. For this reason, optimization methods are widely used for applications and development of the DEA models. Recently, a lot of models have appeared in the literature that expand capabilities of the DEA technologies. Moreover, the notion of selective convexity was introduced in the DEA models. This technology allows managers to consider problems in which there are variables such as averages, ratios, percentages, etc. In this paper, constructions of production functions are considered in DEA models with selective convexity. Computational experiments with real-life datasets confirmed that the proposed algorithms work efficiently. © 2024, Institute of Applied Mathematics of Baku State University. All rights reserved.