arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2310.03095cs.SI
收藏

通过非合作微分博弈实现意见动态优化

Opinion Dynamics Optimization Through Noncooperative Differential Games

Hossein B. Jond

AI总结:

本文研究基于Hegselmann-Krause模型的社交网络意见动态优化,提出非合作微分博弈模型,利用Pontryagin原理推导Nash均衡策略,并在Zachary空手道俱乐部网络上验证发现意见趋近均值但未必达成共识。

摘要:

本文研究在图上优化由个体组成的社会网络的意见形成,其意见根据Hegselmann-Krause意见动态模型演化。我针对非固执的个体群体,提出了一个基于微分博弈的优化问题。每个个体的目标是通过优化个人性能指标,为自己的意见演化寻求最优控制策略。利用Pontryagin极大值原理,推导了该意见优化模型的Nash均衡行动及与均衡行动相关的意见轨迹。博弈策略在著名的Zachary空手道俱乐部社交网络上执行。与博弈策略相关的最终意见轨迹表明,在非固执的Zachary网络中,意见向网络的平均意见靠拢,但最终意见的共识并不必然出现。

英文摘要:

In this paper, I study optimizing the opinion formation of a social network of a population of individuals on a graph whose opinion evolves according to the Hegselmann-Krause model for opinion dynamics. I propose an optimization problem based on a differential game for a population of individuals who are not stubborn. The objective of each individual is to seek an optimal control policy for her own opinion evolution by optimizing a personal performance index. The Nash equilibrium actions and the associated opinion trajectory with the equilibrium actions are derived for the opinion optimization model using Pontryagin's principle. The game strategies were executed on the well-known Zachary's Karate Club social network. The resulting opinion trajectories associated with the game strategies showed that in non-stubborn Zachary's network, the opinions moved toward the average opinion of the network, but a consensus of final opinions did not necessarily emerge.

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。

↑