Introducing a Relational Network DEA Model with Stochastic Intermediate measures for Portfolio Optimization

Document Type : Research Paper


Department of Applied Mathematics, Yadegar-e-Imam Khomeini (RAH) Shahr-e-rey Branch, Islamic Azad University, Tehran, Iran


Conflict intermediate measures in DEA models, especially in constraint and open the black box, is the main difference between traditional DEA and network DEA models. Furthermore, from the application's perspective, intermediate measures aren’t deterministic. So, for measuring the efficiency more precisely, they can be considered as imprecise data. The aim of this paper is introducing a stochastic relational model for measuring overall efficiency that deals with intermediate and outputs as stochastic data. The proposed model is applied for portfolio optimization. An actual data set of 27 Iranian stock industries is applied as numerical example. The result shows that SR-NDEA has better discriminant power than R-NDEA model.


[1] Artzner et al. (1999). Coherent measures of risk. Mathematical Finance, 9(3) (1999), 203– 228.
[2] Behzadi, M. H, Nematollahi, N. Mirbolouki, M. Ranking Efficient DMUs with Stochastic Data by Considering Inecient Frontier, International Journal of Industrial Mathematics, 1 (3): (2009) 219-226.
[3] Charnes A., W.W. Cooper and E. Rhodes. Measuring the Efficiencies of DMUs, European Journal of Operational Research,( 1978) 2(6)429-444.