Article Contents
Article Contents

# Network data envelopment analysis with fuzzy non-discretionary factors

• * Corresponding author: C.-F. Hu
• Network data envelopment analysis (DEA) concerns using the DEA technique to measure the relative efficiency of a system, taking into account its internal structure. The results are more meaningful and informative than those obtained from the conventional DEA models. This work proposed a new network DEA model based on the fuzzy concept even though the inputs and outputs data are crisp numbers. The model is then extended to investigate the network DEA with fuzzy non-discretionary variables. An illustrative application assessing the impact of information technology (IT) on firm performance is included. The results reveal that modeling the IT budget as a fuzzy non-discretionary factor improves the system performance of firms in a banking industry.

Mathematics Subject Classification: Primary: 90B50, 90C08; Secondary: 91B06.

 Citation:

• Figure 1.  General network systems [12]

Figure 2.  Network system discussed in [18]

Table 1.  Data set for assessing IT impact on firm performance

 DMU j IT Fixed No. of Deposits Profit Fraction $\rm {budget}$ ${\mbox{assets}}$ ${\mbox{employees }}$ of loans $({＄ \ \mbox{billions})}$ $({＄ \ \mbox{billions})}$ $({＄ \ \mbox{billions})}$ $({＄ \ \mbox{billions})}$ $({＄ \ \mbox{billions})}$ ${\mbox{recovered}}$ $X_1$ $X_2$ $X_3$ $Z$ $Y_1$ $Y_2$ 1 $0.150$ $0.713$ $13.3$ $14.478$ $0.232$ $0.986$ 2 $0.170$ $1.071$ $16.9$ $19.502$ $0.340$ $0.986$ 3 $0.235$ $1.224$ $24.0$ $20.952$ $0.363$ $0.986$ 4 $0.211$ $0.363$ $15.6$ $13.902$ $0.211$ $0.982$ 5 $0.133$ $0.409$ $18.485$ $15.206$ $0.237$ $0.984$ 6 $0.497$ $5.846$ $56.42$ $81.186$ $1.103$ $0.955$ 7 $0.060$ $0.918$ $56.42$ $81.186$ $1.103$ $0.986$ 8 $0.071$ $1.235$ $12.0$ $11.441$ $0.199$ $0.985$ 9 $1.500$ $18.120$ $89.51$ $124.072$ $1.858$ $0.972$ 10 $0.120$ $1.821$ $19.8$ $17.425$ $0.274$ $0.983$ 11 $0.120$ $1.915$ $19.8$ $17.425$ $0.274$ $0.983$ 12 $0.050$ $0.874$ $13.1$ $14.342$ $0.177$ $0.985$ 13 $0.370$ $6.918$ $12.5$ $32.491$ $0.648$ $0.945$ 14 $0.440$ $4.432$ $41.9$ $47.653$ $0.639$ $0.979$ 15 $0.431$ $4.504$ $41.1$ $52.63$ $0.741$ $0.981$ 16 $0.110$ $1.241$ $14.4$ $17.493$ $0.243$ $0.988$ 17 $0.053$ $0.450$ $7.6$ $9.512$ $0.067$ $0.980$ 18 $0.345$ $5.892$ $15.5$ $42.469$ $1.002$ $0.948$ 19 $0.128$ $0.973$ $12.6$ $18.987$ $0.243$ $0.985$ 20 $0.055$ $0.444$ $5.9$ $7.546$ $0.153$ $0.987$ 21 $0.057$ $0.508$ $5.7$ $7.595$ $0.123$ $0.987$ 22 $0.098$ $0.370$ $14.1$ $16.906$ $0.233$ $0.981$ 23 $0.104$ $0.395$ $14.6$ $17.264$ $0.263$ $0.983$ 24 $0.206$ $2.680$ $19.6$ $36.430$ $0.601$ $0.982$ 25 $0.067$ $0.781$ $10.5$ $11.581$ $0.120$ $0.987$ 26 $0.100$ $0.872$ $12.1$ $22.207$ $0.248$ $0.972$ 27 $0.0106$ $1.757$ $12.7$ $20.670$ $0.253$ $0.988$

Table 2.  The system efficiency, $\theta_p^{\ast},$ and the membership degree, $\alpha_p, p = 1, 2, \cdots, 27.$

 DMU j Model (2) $\theta^{\ast}$ ${ \text{Model (6)}}$ DMU j Model (2)$\theta^{\ast}$ ${ \text{Model (6)}}$ $\alpha^{\ast}$ $1-\alpha^{\ast}$ $\alpha^{\ast}$ $1-\alpha^{\ast}$ $1$ $0.6388$ $0.3612$ $0.6388$ $15$ $0.6582$ $0.3418$ $0.6582$ $2$ $0.6507$ $0.3493$ $0.6507$ $16$ $0.6646$ $0.3354$ $0.6646$ $3$ $0.5179$ $0.4821$ $0.5179$ $17$ $0.7177$ $0.2823$ $0.7177$ $4$ $0.5986$ $0.4014$ $0.5986$ $18$ $1.0000$ $0.0000$ $1.0000$ $5$ $0.5556$ $0.4444$ $0.5556$ $19$ $0.8144$ $0.1856$ $0.8144$ $6$ $0.7599$ $0.2401$ $0.7599$ $20$ $0.6940$ $0.3060$ $0.6940$ $7$ $1.0000$ $0.0000$ $1.0000$ $21$ $0.7067$ $0.2933$ $0.7067$ $8$ $0.5352$ $0.4648$ $0.5352$ $22$ $0.7942$ $0.2058$ $0.7942$ $9$ $0.6249$ $0.3751$ $0.6249$ $23$ $0.7802$ $0.2198$ $0.7802$ $10$ $0.4961$ $0.5039$ $0.4961$ $24$ $0.9300$ $0.0700$ $0.9300$ $11$ $0.4945$ $0.5055$ $0.4945$ $25$ $0.6270$ $0.3730$ $0.6270$ $12$ $0.6685$ $0.3315$ $0.6685$ $26$ $1.0000$ $0.0000$ $1.0000$ $13$ $0.9487$ $0.0513$ $0.9487$ $27$ $1.0000$ $0.0000$ $1.0000$ $14$ $0.5880$ $0.4120$ $0.5880$

Table 3.  The results of solving the proposed fuzzy non-discretionary Model (14)

 $\begin{array}{c} \mbox{DMU}\\ j \end{array}$ Fuzzy non-discretionary input $\bar{X}_{1j}^{\ast}$ $\bar{X}_{2j}^{\ast}$ $\bar{X}_{3j}^{\ast}$ $\alpha^{\ast}$ $1-\alpha^{\ast}$ Rank 1 0.1102 0.5236 9.6335 0.2654 0.7346 18 2 0.1260 0.7723 12.4259 0.2589 0.7411 17 3 0.1586 0.8079 16.1328 0.3253 0.6747 25 4 0.1506 0.2564 10.9013 0.2864 0.7136 21 5 0.0921 0.2793 11.2165 0.3077 0.6923 23 6 0.4008 4.6342 45.4289 0.1936 0.8064 10 7 0.0600 0.9180 56.4200 0.0000 1.0000 1 8 0.0485 0.7677 8.0988 0.3173 0.6827 24 9 1.0908 13.1471 64.8529 0.2728 0.7272 20 10 0.0798 1.0715 12.8295 0.3351 0.6649 26 11 0.0797 1.0997 12.7416 0.3358 0.6642 27 12 0.0376 0.6544 9.7883 0.2490 0.7510 14 13 0.3519 5.3291 11.8900 0.0488 0.9512 5 14 0.3116 3.1047 29.5929 0.2918 0.7082 22 15 0.3212 3.2772 30.4969 0.2547 0.7453 16 16 0.0824 0.8943 10.7729 0.2512 0.7488 15 17 0.0413 0.3509 5.8871 0.2202 0.7798 11 18 0.3450 5.8920 15.5000 0.0000 1.0000 1 19 0.1080 0.8154 10.6151 0.1565 0.8435 7 20 0.0421 0.3349 4.4948 0.2343 0.7657 13 21 0.0441 0.3904 4.3686 0.2268 0.7732 12 22 0.0813 0.3043 11.6775 0.1707 0.8293 8 23 0.0853 0.3216 11.9554 0.1802 0.8198 9 24 0.1925 2.4125 18.3176 0.0654 0.9346 6 25 0.0488 0.5448 7.5342 0.2717 0.7283 19 26 0.1000 0.8720 12.1000 0.0000 1.0000 1 27 0.0106 1.7570 12.7000 0.0000 1.0000 1
•  [1] R. D. Banker and R. Morey, Efficiency analysis for exogenously fixed inputs and outputs, Oper. Res., 34 (1986), 501-653.  doi: 10.1287/opre.34.4.513. [2] M. Barat, G. Tohidi and M. Sanei, DEA for nonhomogeneous mixed networks, Asia Pac. Manag. Rev., 24 (2018), 161-166.  doi: 10.1016/j.apmrv.2018.02.003. [3] R. E. Bellman and L. A. Zadeh, Decision making in a fuzzy environment, Manag. Sci., 17 (1970), B141–B164. doi: 10.1287/mnsc.17.4.B141. [4] L. Castelli, R. Pesenti and W. Ukovich, DEA-like models for the efficiency evaluation of hierarchically structured units, Eur. J. Oper. Res., 154 (2004), 465-476.  doi: 10.1016/S0377-2217(03)00182-6. [5] J. Zhu, Data Envelopment Analysis: A Handbook of Modeling Internal Structures and Networks, International Series in Operations Research & Management Science, 238. Springer, New York, 2016. doi: 10.1007/978-1-4899-7684-0. [6] J. M. Cordero-Ferrera, F. Pedraja-Chaparro and D. Santín-González, Enhancing the inclusion of non-discretionary inputs in DEA, J. Oper. Res. Soc., 61 (2010), 574-584.  doi: 10.1057/jors.2008.189. [7] R. Färe and S. Grosskopf, Intertemporal Production Frontiers: With Dynamic DEA, Boston: Kluwer Academic Publishers, 1996. [8] R. Färe and S. Grosskopf, Network DEA, Socio. Econ. Plann. Sci., 4 (2000), 35-49. [9] D. U. A. Galagedera, Modelling social responsibility in mutual fund performance appraisal: A two-stage data envelopment analysis model with non-discretionary first stage output, Eur. J. Oper. Res., 273 (2019), 376-389.  doi: 10.1016/j.ejor.2018.08.011. [10] B. Golany and Y. Roll, Some extensions of techniques to handle non-discretionary factors in data envelopment analysis, J. Prod. Anal., 4 (1993), 419-432.  doi: 10.1007/BF01073549. [11] C. Kao, Network data envelopment analysis: A review, Eur. J. Oper. Res., 239 (2014), 1-16.  doi: 10.1016/j.ejor.2014.02.039. [12] C. Kao, Efficiency decomposition and aggregation in network data envelopment analysis, Eur. J. Oper. Res., 255 (2016), 778-786.  doi: 10.1016/j.ejor.2016.05.019. [13] C. Kao and S.-N. Hwang, Efficiency measurement for network systems: IT impact on firm performance, Decis. Support Syst., 48 (2010), 437-446.  doi: 10.1016/j.dss.2009.06.002. [14] R. J. Kauffman and P. Weill, An evaluative framework for research on the performance effects of information technology investment, Proceedings of the 10th International Conference on Information Systems, (1989), 377–388. doi: 10.1145/75034.75066. [15] M. A. Muniz, J. Paradi, J. Ruggiero and Z. Yang, Evaluating alternative DEA models used to control for non-discretionary inputs, Comput. Oper. Res., 33 (2006), 1173-1183. [16] L. Simar and P. W. Wilson, Estimation and inference in two-stage, semi-parametric models of production processes, J. Econom., 136 (1997), 31-64.  doi: 10.1016/j.jeconom.2005.07.009. [17] M. Taleb, R. Ramli and R. Khalid, Developing a two-stage approach of super efficiency slack-based measure in the presence of non-discretionary factors and mixed integer-valued data envelopment analysis, Expert. Syst. Appl., 103 (2018), 14-24.  doi: 10.1016/j.eswa.2018.02.037. [18] C. H. Wang, R. Gopal and S. Zionts, Use of data envelopment analysis in assessing information technology impact on firm performance, Ann. Oper. Res., 73 (1997), 191-213. [19] M. Zerafat Angiz L and A. Mustafa, Fuzzy interpretation of efficiency in data envelopment analysis and its application in a non-discretionary model, Knowl.-Based Syst., 49 (2013), 145-151.

Figures(2)

Tables(3)