Does Fuzzy AHP Really Adds Value?

What is it about?

both AHP and its extension, Fuzzy AHP (FAHP), have been extensively studied and applied to a variety of MCDM problems. Given that AHP relies on subjective judgments from decision-makers to determine criteria weights and alternative scores, and that these judgments are often imprecise, it is natural to model them using fuzzy sets rather than exact numbers. This viewpoint has prompted the development of a variety of FAHP algorithms in literature. But do they outperform AHP? In this paper we search the answer to this question. Our findings is somewhat discouraging for many researchers.

Featured Image

Why is it important?

We scrutinize the benefits of incorporating fuzziness into conventional AHP. To address this, we developed a comprehensive framework and set of metrics to evaluate and compare the performance of FAHP algorithms. We conducted extensive comparisons between fuzzy and conventional AHP models using Monte Carlo simulations and empirical data. Our analysis revealed that, despite the intuitive appeal of using fuzzy numbers instead of crisp ones to represent the linguistic judgments provided by decision-makers, the existing FAHP algorithms were actually outperformed by traditional AHP methods.

Perspectives