Hybrid models in decision making under uncertainty: The case of training provider evaluation

This paper provides a novel design for two hybrid models in modeling decision making under uncertainty: AHP-Fuzzy PROMETHEE and AHP-Fuzzy TOPSIS. The analytic hierarchy process' (AHP) excellent ability in problem structuring allows weights of criteria to be easily gathered from experts in the decision problem. Nonetheless, the pairwise comparisons required are immense, thus inducing decision making fatigue as the number of evaluation objects and criteria increase. We show that the number of pairwise comparisons can be reduced by integrating PROMETHEE or TOPSIS into AHP. The former two techniques are distance based methods. PROMETHEE allows the evaluators to choose a set of preference function and calculates the distance between the evaluator’s judgment and his limits. TOPSIS, on the other hand, computes the distance of a judgment from the best and worst cases. Fuzzy linguistics are incorporated into PROMETHEE and TOPSIS, thus effectively modeling decision making subjectivity – aside from eliminating the need for evaluators to specify their preference limits in PROMETHEE. These techniques are applied in a strategic outsourcing decision of a company that seeks to evaluate their training providers. The final results indicate that both AHP-Fuzzy TOPSIS and AHP-Fuzzy PROMETHEE achieved consistent results and arrived at the same ranking order.