Stability of Surrogate MCDA Weights Under Different Assumptions on Value Distributions

Lakmayer, S. & Danielson, M. (2025). Stability of Surrogate MCDA Weights Under Different Assumptions on Value Distributions. In: Human and Artificial Intelligence in Group Decision and Negotiation. Eds. Moreno-Jiménez, J.M., Costa Morais, D., Escobar, M.T., & Turón, A., pp. 68-81 Cham: Springer Nature Switzerland. ISBN 978-3-031-95221-0 10.1007/978-3-031-95221-0_5.

Full text not available from this repository.

Abstract

This article studies the effects of sampling the alternative values from different distributions when studying the performance of surrogate weight methods in the additive model in multi-criteria decision analysis. The aim is to demonstrate surrogate weights' performance invariance regarding underlying actual distributions. Multiple distributions are characterised and examined through extensive simulation to evaluate their influence on the efficacy of surrogate weight approaches. It was found that employing the presently accepted standard distributions for alternative values led to outcomes from the surrogate weight methods that were remarkably consistent, barring a few notable deviations—-suggesting considerable robustness. In contrast, drawing samples from more extreme distributions resulted in greater divergence. Overall, the observed patterns of the surrogate weight approaches align, to a substantial degree, with findings reported in prior research. We conclude that the performance of the surrogate weight methods is generally stable under a wide variety of reasonable alternative value distributions and show a case when the distribution is too skewed.

Item Type:	Book Section
Research Programs:	Advancing Systems Analysis (ASA) Advancing Systems Analysis (ASA) > Cooperation and Transformative Governance (CAT)
Depositing User:	Luke Kirwan
Date Deposited:	25 Jun 2025 07:18
Last Modified:	25 Jun 2025 07:18
URI:	https://pure.iiasa.ac.at/20709

Actions (login required)

View Item