Applications of efficiency and complex network analysis to logistics and dominance networks

Abstract

The major purpose of this Thesis is to contribute to the knowledge of Data Envelopment Analysis (DEA). Data Envelopment Analysis is a technique that aims to evaluate the efficiency of homogeneous productive units and to determine benchmarks to guide inefficient units to achieve the Efficient Frontier. Several contributions to DEA are sought in this Thesis. One of the main goals of this Thesis is to develop a new approach to analyse the dominance relationships that appear between the assessed units in a DEA context. This approach is a methodology based on Complex Network Analysis. Complex Network Analysis (CNA) is a traditional technique to characterise the relationships among the elements of a system. It has a powerful visualisation tool that can be applied to a DEA context. The first methodological approach introduced here is called Dominance Network Analysis (DNA). Since it is based on CNA, it is flexible enough to represent the different relationships usually studied in DEA. In DEA there are many different models designed to evaluate the relative technical efficiency, but also the economic efficiency. The proposed DNA metholodogy is capable of evaluating both relationships at the same time, as described in three of the papers included in this Thesis, using different datasets to show the potential of this new approach. Another main goal of this Thesis is to develop a new concept to determine intermediate benchmarks that the inefficient units can achieve in the technology seeking the Efficient Frontier. The new concept, called Gradient-based stepwise benchmarking paths, is based on the Efficiency Field Potential (EFP). This approach determines consecutive intermediate benchmarks, depending on the position of the evaluated Decision Making Unit (DMU). The last contribution of this Thesis is to explore the potential of DEA in a project team assessment context. In teamwork, one of the most difficult problems is to measure the contribution of each member to the team. However, if this contribution were known, there should exist a relationship between the efficiency of the team members and the efficiency of their developed projects. So, the efficiency of each team member could be estimated based on the efficiency of their results.