集体决策的背包问题
A knapsack for collective decision-making
AI总结:
本研究通过结合偏好聚合与背包分配方法,探讨如何高效汇总不同利益相关者的投入以选择最优项目组合,发现多群体、适度专业差异及专家委托可提升长期价值识别能力。
摘要:
集体决策是不同利益相关者达成联合决策的过程。在社会环境中,一个例子是参与式预算,即由 constituents 决定公共项目的资金分配。如何最有效地汇总不同利益相关者对具有不确定长期收益的项目组合的投入,仍然是一个未解决的问题。我们通过将偏好聚合与背包分配方法相结合来研究集体决策,从而解决这一问题。由于不同的利益相关者群体可能对项目有不同的评估,我们考察了几种聚合方法,以结合他们的多样化投入。聚合后的评估随后用于填充一个“集体”背包。在我们考虑的方法中,包括算术平均、Borda型排名以及委托给专家。我们发现,提高聚合方法识别具有最大预期长期价值项目的能力的因素包括:拥有多个利益相关者群体、其专业水平存在适度差异,以及一定程度的委托或偏向于那些更能客观评估项目的群体。我们还讨论了评估错误和异质性成本如何影响项目选择。我们提出的聚合方法不仅适用于公共项目资助的背景,而且更普遍地适用于不确定性下的组织决策。
英文摘要:
Collective decision-making is the process through which diverse stakeholders reach a joint decision. Within societal settings, one example is participatory budgeting, where constituents decide on the funding of public projects. How to most efficiently aggregate diverse stakeholder inputs on a portfolio of projects with uncertain long-term benefits remains an open question. We address this problem by studying collective decision-making through the integration of preference aggregation and knapsack allocation methods. Since different stakeholder groups may evaluate projects differently,we examine several aggregation methods that combine their diverse inputs. The aggregated evaluations are then used to fill a ``collective'' knapsack. Among the methods we consider are the arithmetic mean, Borda-type rankings, and delegation to experts. We find that the factors improving an aggregation method's ability to identify projects with the greatest expected long-term value include having many stakeholder groups, moderate variation in their expertise levels, and some degree of delegation or bias favoring groups better positioned to objectively assess the projects. We also discuss how evaluation errors and heterogeneous costs impact project selection. Our proposed aggregation methods are relevant not only in the context of funding public projects but also, more generally, for organizational decision-making under uncertainty.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。