Stochastic Approaches for Criteria Weight Identification in Multi-criteria Decision Analysis

This paper delves into the significance of determining criteria weights, a pivotal aspect influencing outcomes in Multi-Criteria Decision Analysis (MCDA) methods. While traditional approaches often rely on information measures or expert knowledge, this study introduces an innovative Stochastic IdenTification of Weights (SITW) method. Leveraging the Particle Swarm Optimization (PSO) algorithm, SITW integrates with the Characteristic Objects Method (COMET). By employing the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), the PSO method explores weight spaces dynamically. This integration enhances the model’s adaptability and resilience and allows for the identification of global weights closely aligning with local weights in the reference model. The study highlights the potential of SITW in addressing uncertainties within complex decision scenarios, marking a significant advancement in decision analysis field.