Speaker
Description
Multi-Criteria Decision Making (MCDM) has long emphasized the importance of modeling interactions among criteria to improve the evaluation and ranking of alternatives. Fuzzy measures and the Choquet integral provide a powerful non-additive framework for capturing synergistic and redundant relationships among criteria and have been widely applied in decision analysis. However, the effectiveness of these methods largely depends on the quality of fuzzy measure elicitation. Existing studies have primarily focused on fuzzy measure identification algorithms, parameter estimation techniques, and aggregation model improvements. In contrast, limited attention has been paid to the measurement quality of the elicitation process itself, despite the fact that cognitive and measurement biases may distort respondents’ evaluations and subsequently affect fuzzy measure estimation and fuzzy integral outcomes. Consequently, the potential impact of survey design on fuzzy measure quality remains insufficiently explored. To address this gap, this study develops two enhanced fuzzy measure elicitation approaches, namely a Slider-Based Measurement method and a Semantic-Based Measurement method. By comparing the fuzzy measures and Choquet integral results generated by traditional and proposed approaches, this study investigates potential measurement bias in conventional elicitation procedures. Furthermore, a predictive validity framework is employed to evaluate whether the proposed methods more accurately capture respondents’ underlying preferences and cognitive judgments.
| Keywords | Fuzzy measures, Choquet integral, Non-additive, The quality of fuzzy measure elicitation, A predictive validity framework |
|---|---|
| Participation in the SNSF Early-Stage Researcher Support and Award Scheme | I am submitting a regular presentation and would like to participate in the SNSF Early-Stage Researcher Support and Award Scheme. |