Clustering Based on Data Envelopment Analysis: Application to Management Research and Practice  
Author Valentina Kuskova

 

Co-Author(s) Dmitry Zaytsev

 

Abstract Data Envelopment Analysis (DEA), a linear programming technique, is becoming increasingly popular in management studies. A variety of extensions have been proposed since its first introduction in the late 1970s, and lately, the idea of using the tool for clustering data has received some attention in the literature. We have further extended the idea of DEA-based clustering by introducing the notion of clustering by DEA-generated vector, with some useful applications in non-homogeneous samples, data with unclear theoretical structure, or as an explanatory tool. We demonstrate how these techniques can be used in management on a sample from education industry.

 

Keywords Data envelopment analysis, clustering, performance measurement
   
    Article #:  RQD25-142
 
Proceedings of 25th ISSAT International Conference on Reliability & Quality in Design
August 1-3, 2019 - Las Vegas, NV, U.S.A.