15.5 Probabilistic Additive Weighting
15.5.1 Introduction
Variations in results from a decision model are usually a consequence of the
variability associated with estimates of the various criteria under consideration. As
shown in Sect. 15.4, a sensitivity analysis is one way to evaluate the effects of such
changes in criteria valuations on the final result. Another method for evaluating the
effect of alterations in criterion scores on the final result is to directly incorporate the
uncertainty associated with the criterion estimates into the additive weighting model.
The two basic probability concepts required within this analysis are expected value
and variance. Let us examine these two briefly.
Ranking
Baseline ST1
ST2
ST3
ST4
ST5
ST6
1
2
3
4
5
6
(a) Two best performing options
Site 1
Site 4
Ranking
Baseline ST1
ST2
ST3
ST4
ST5
ST6
1
2
3
4
5
6
(b) Other four options
Site 3
Site 5
Site 6
Site 2
Fig. 15.1 Baseline and sensitivity test results
422
15 Multi Objective Evaluation Criteria for Infrastructure
15.5.1 Introduction
Variations in results from a decision model are usually a consequence of the
variability associated with estimates of the various criteria under consideration. As
shown in Sect. 15.4, a sensitivity analysis is one way to evaluate the effects of such
changes in criteria valuations on the final result. Another method for evaluating the
effect of alterations in criterion scores on the final result is to directly incorporate the
uncertainty associated with the criterion estimates into the additive weighting model.
The two basic probability concepts required within this analysis are expected value
and variance. Let us examine these two briefly.
Ranking
Baseline ST1
ST2
ST3
ST4
ST5
ST6
1
2
3
4
5
6
(a) Two best performing options
Site 1
Site 4
Ranking
Baseline ST1
ST2
ST3
ST4
ST5
ST6
1
2
3
4
5
6
(b) Other four options
Site 3
Site 5
Site 6
Site 2
Fig. 15.1 Baseline and sensitivity test results
422
15 Multi Objective Evaluation Criteria for Infrastructure