The major advantage of DEA is to reduce multivariate data to a single key performance indicator (KPI). However, this KPI is just one element to determine the (economic) situation of a Decision-Making Unit (DMU). Beside this, DEA offers more information about the DMU under consideration; more precisely: optimal multiplier weights, reference sets, returns to scale, etc. The purpose of this contribution is to identify similarities or dissimilarities within the set of DMUs in a dynamic context, applying the aforementioned DEA results. Therefore, Multidimensional Scaling will be used. But here, the question arises: what data should be selected – regarding the decision situation – at best?
In this talk we focus on this question by investigating different combinations of data, i.e., the decomposition of efficiency scores, raw data of DMUs (activities) as well as datasets joining exogenous and endogenous variables. The results are illustrated by numerical examples.
Keywords: Multidimensional scaling, visualization, decision support, datasets, dynamic analysis
The purpose of a peer-based DEA is to find a peer – one price system that acts
as a common efficiency- and RTS-denominator for all DMUs. The peer price
system then might be used for further activity planning. In most applications
the price system with the least efficiency deviation for all DMUs is the preferred
one. Efficiency is only one economic measure, RTS is the other.
In this talk we show that efficiencies and returns to scale are both appropriate
indices to select a peer. However, in some cases these objectives may collide
and hence we get different peers. Therefore, we present a multi-criteria
optimization problem to combine both potentially conflicting criteria.
Keywords: BCC-efficiency -- returns to scale -- consensual peer -- multi-objective optimization