| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 229.74 KB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
The additive model of multiattribute value (or utility) the
ory is widely
used in multicriteria choice problems. However, often it is
not easy to
obtain precise values for the scaling weights or the alterna
tives’ value in
each function. Several decision rules have been proposed to
select an
alternative under these circumstances, which require weak
er information,
such as ordinal information. We propose new decision rules a
nd test
them using Monte-Carlo simulation, considering that there
exists ordinal
information both on the scaling weights and on the alternati
ves’ values.
Results show the new rules constitute a good approximation.
We provide
guidelines about how to use these rules in a context of select
ing a subset of
the most promising alternatives, considering the contradi
ctory objectives
of keeping a low number of alternatives yet not excluding the
best one.
Description
Keywords
Multi-Criteria Decision Analysis MAUT/MAVT Imprecise/ incom- plete/ partial information Ordinal information Simulat ion