Abstract:As decision-making scenarios become increasingly complex, this paper introduces a novel approach to stochastic multi-attribute decision-making (SMADM) through a distributionally robust optimization lens. The core of this study lies in the development of a Chance-Constrained SMADM (CC-SMADM) model that effectively handles uncertainties in evaluation values without the prerequisite of known probability density functions. This research makes a significant leap by employing the Wasserstein measure to construct an ambiguity set, which serves as a proxy for the true but unknown distribution of random variables. The model reformulates distributionally robust chance constraints into a computationally tractable form, utilizing the worst-case Conditional Value at Risk (CVaR). This work presents a conic optimization problem solvable with standard software, marking a substantial advancement in the field of decision-making under uncertainty. The proposed approach not only increases the robustness of decision-making models but also broadens their applicability across various scenarios where the distribution of attributes is either complex, unknown, or non-standard.

A Distributionally Robust Optimization Approach for Chance-Constrained Stochastic Multi-Attribute Decision-Making with Unknown Probability Density Function