The Comparison of Principal Component Analysis and Data Envelopment Analysis in Ranking of Decision Making Units

Authors : Filiz KARDIYEN, Hhasan ÖRKCÜ

Pages : 127-133

View : 24 | Download : 9

Publication Date : 2010-03-29

Article Type : Other Papers

Abstract :Normal 0 21 false false false TR X-NONE X-NONE In this study, Data Envelopment Analysis insert ignore into journalissuearticles values(DEA); and Principal Component Analysis insert ignore into journalissuearticles values(PCA); were compared when these two methods are used for ranking Decision Making Units insert ignore into journalissuearticles values(DMU); with multiple inputs and outputs. DEA, a nonstatistical technique, is a methodology using a linear programming model for evaluating and ranking DMU’s performance. PCA, a multivariate statistical method, uses new measures defined by DMU’s inputs and outputs. The results of both methods were applied to a real data set that indicates the economic performances of European Union member countries and also, a simulation study was done for different sample sizes and for different numbers of input-output, and the results were examined. For both applications, consistent results were obtained. Spearman’s correlation test is employed to compare the rankings obtained by PCA and DEA.   Key Words: Principal Component Analysis, Data Envelopment Analysis, Ranking  
Keywords :