An Experimental Comparison of Different Rules for Aggregation, in AHP

摘要

The current study presents a comparative analysis of different aggregation rules (for instance, two traditional aggregation rules; additive & multiplicative and its new variants the exponential & logarithmic transformation) adopted in Analytic Hierarchy Process (AHP), by testing them against a problem with a known composite answer. The result showed the additive aggregation rule seems to be more satisfactory than the others. In continuation, we generated 100 decision matrixes with random numbers in the range of 1 to 9 (in consistent to AHP scale). The result indicated that, in the both introduced new aggregation rules (exponential & logarithmic transformation) the portion of important alternative values come to other alternatives. Particularly, to worst alternatives. Therefore, in these transformations, discriminating level of the most important alternative will decrease.