随机博弈
旅行者困境
社会困境
困境
机制(生物学)
选择(遗传算法)
价值(数学)
计算机科学
数理经济学
囚徒困境
微观经济学
经济
博弈论
人工智能
数学
正常形式游戏
重复博弈
机器学习
认识论
哲学
几何学
作者
Xinle Lin,Jianhe Li,Suohai Fan
出处
期刊:Chaos
[American Institute of Physics]
日期:2024-08-01
卷期号:34 (8)
被引量:1
摘要
We proposed a neighbor selection mechanism based on memory and target payoff, where the target payoff is the maximum value of the group’s average expected payoff. According to this mechanism, individuals prioritize selecting neighbors whose average payoffs in the last M rounds are close to the target payoff for strategy learning, aiming to maximize the group’s expected payoff. Simulation results on the grid-based Prisoner’s Dilemma and Snowdrift games demonstrate that this mechanism can significantly improve the group’s payoff and cooperation level. Furthermore, the longer the memory length, the higher the group’s payoff and cooperation level. Overall, the combination of memory and target payoff can lead to the emergence and persistence of cooperation in social dilemmas as individuals are motivated to cooperate based on both their past experiences and future goals. This interplay highlights the significance of taking into account numerous variables in comprehending and promoting cooperation within evolutionary frameworks.
科研通智能强力驱动
Strongly Powered by AbleSci AI