Tagungsvortrag

Titel:

Exploratory benchmarking with DEA and multidimensional scaling - process, reference model and application

AutorInnen:

Steffen Hoffmann

Kategorie:

Wissenschaftliche Vorträge

Forschungsthema:

Data Envelopment Analysis

 

33rd European conference on Operational Research, 1. Juli 2024, Kopenhagen.

Abstract:

Benchmarking is a valuable process for improving strategic and operational decisions in both profit and non-profit organizations. An important sub-process of benchmarking is to measure performance gaps and uncover their causes. While Data Envelopment Analysis (DEA) can handle the evaluation part, Data Mining tasks are suitable to explore the link between computational results and additional knowledge. Therefore, a new Data Mining process called Exploratory Benchmarking is proposed. The process integrates DEA and Multidimensional Scaling (MDS) as a reference model to structure decision support using wellresearched methodologies. First, the role of DEA is to evaluate the related data in context. Next, MDS is used to visualize the results of the evaluation. Finally, a systematic analysis is performed to uncover manifest and latent structures using a specially developed exploration procedure. The entire process is exemplified by a real-world application in the field of air navigation service providers at airports. However, the approach can be applied far beyond this as a standard process in business analysis and consulting.