標題: A network-based approach for increasing discrimination in data envelopment analysis
作者: Liu, J. S.
Lu, W-M
Yang, C.
Chuang, M.
交大名義發表
National Chiao Tung University
關鍵字: data envelopment analysis;linear programming;social network analysis;Bonacich centrality
公開日期: 1-十一月-2009
摘要: Data envelopment analysis (DEA) is known to produce more than one efficient decision-making unit (DMU). This paper proposes a network-based approach for further increasing discrimination among these efficient DMUs. The approach treats the system under study as a directed and weighted network in which nodes represent DMUs and the direction and strength of the links represent the relative relationship among DMUs. In constructing the network, the observed node is set to point to its referent DMUs as suggested by DEA. The corresponding lambda values for these referent DMUs are taken as the strength of the network link. The network is weaved by not only the full input/output model, but also by models of all possible input/output combinations. Incorporating these models into the system basically introduces the merits of each DMU under various situations into the system and thus provides the key information for further discrimination. Once the network is constructed, the centrality concept commonly used in social network analysis-specifically, eigenvector centrality-is employed to rank the efficient DMUs. The network-based approach tends to rank high the DMUs that are not specialized and have balanced strengths.
URI: http://dx.doi.org/10.1057/jors.2009.35
http://hdl.handle.net/11536/6465
ISSN: 0160-5682
DOI: 10.1057/jors.2009.35
期刊: JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY
Volume: 60
Issue: 11
起始頁: 1502
結束頁: 1510
顯示於類別:期刊論文