Expert opinions aggregation is a generic part of the group decision making (GDM) problem. The challenge of expert opinions aggregation is to reduce the subjectivity in the process as much as possible and improve the reliability of the aggregated opinion. Most of the existing literature try to eliminate the subjectivity but seldom consider the reliability of the aggregation result. In this article, we propose a new criterion that contains consensus level and confidence level to improve both objectivity (i.e., consensus) and reliability (i.e., no absurd result) with the experts’ opinions being represented as probability density functions. Subsequently, the expert opinion aggregation problem is formulated as a biobjective optimization model. The Survey of Professional Forecasters is used as an example to examine the feasibility and accuracy of the proposed approach and the result shows that the new approach can provide a better estimation than that of the single objective model in the literature. To our best knowledge, the proposed criterion is new in the literature of GDM along with relevant problems. The proposed criterion is actually a pilot work to probe the problem of the quality of a GDM process, which is largely ignored in the field of GDM.

中文翻译：

---

专家意见聚合的双目标优化模型及其在群体决策中的应用

专家意见聚合是群体决策 (GDM) 问题的通用部分。专家意见聚合的挑战在于尽可能减少过程中的主观性，提高聚合意见的可靠性。现有文献大多试图消除主观性，但很少考虑聚合结果的可靠性。在本文中，我们提出了一个新标准，其中包含共识水平 和 置信度提高客观性（即共识）和可靠性（即没有荒谬的结果），将专家的意见表示为概率密度函数。随后，专家意见聚合问题被表述为双目标优化模型。以专业预报员调查为例，检验了所提出方法的可行性和准确性，结果表明，与文献中的单一目标模型相比，新方法可以提供更好的估计。据我们所知，所提出的标准是 GDM 文献中的新标准以及相关问题。所提出的标准实际上是探索 GDM 过程质量问题的一项试点工作，这在 GDM 领域中很大程度上被忽视了。