|
Abstract
|
In the Data Envelopment Analysis (DEA) the eciency of the units can be obtained
by identifying the degree of the importance of the criteria (inputs-outputs).In DEA
basic models, challenges are zero and unequal weights of the criteria when decisionmaking
units are evaluated. One of the strategies applied to deal with these problems
is using common weights of the each input/output in all decision making units (DMUs).
In practice the DMUs are containing intermediate process. However, these units are
considered as a black box in DEA basic models, disregarding internal process. This was
the main reason network data envelopment analysis was introduced. On the other hand,
similar challenges mentioned for DEA, zero and unequal weights of the criteria, exist
for network structures as well. This paper suggests a common set of the weights for
network structures to deal with the above problems using nonlinear models, for general
case. Also some numerical examples using proposed models are presented.
|