arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2102.07523cs.MAcs.AI
收藏

基于强化学习的合作和声誉动态

Cooperation and Reputation Dynamics with Reinforcement Learning

University College London(伦敦大学学院);The Alan Turing Institute(艾伦·图灵研究所);Monash University(莫纳什大学);University of Bologna(博洛尼亚大学)

Nicolas Anastassacos, Julian García, Stephen Hailes, Mirco Musolesi

摘要:

在自然和人工系统中,建立合作激励机制是一项挑战。一个潜在的答案是声誉,即代理商交易的直接成本的合作,为未来的利益,有一个良好的声誉。博弈理论模型表明,特定的社会规范可以使合作稳定,但代理人如何能够独立学习,建立有效的声誉机制,对自己的了解较少。我们使用一个简单的强化学习模型来说明声誉机制会产生两个协调问题:代理需要学习如何协调现有声誉的意义,并集体同意一个社会规范,根据他们的行为将声誉分配给其他人。这些协调问题表现出多重均衡,其中一些有效地建立合作。当我们训练代理与标准的Q-学习算法的环境中存在的声誉机制,收敛到不希望的均衡是普遍的。我们提出了两种机制来缓解这一点:(i)用固定的代理人来播种系统的一部分,这些代理人引导其他人走向良好的均衡;以及(ii)基于内省思想的内在奖励,即,增加代理人的奖励的数额成比例的表现,他们自己的战略对自己。这些简单机制的组合在稳定合作方面是成功的,即使在完全分散的问题版本中,代理人同时学习使用和分配声誉。我们展示了我们的研究结果与进化博弈论文献的关系,并讨论了对人工,人类和混合系统的影响,其中声誉可以作为建立信任和合作的一种方式。

英文摘要:

Creating incentives for cooperation is a challenge in natural and artificial systems. One potential answer is reputation, whereby agents trade the immediate cost of cooperation for the future benefits of having a good reputation. Game theoretical models have shown that specific social norms can make cooperation stable, but how agents can independently learn to establish effective reputation mechanisms on their own is less understood. We use a simple model of reinforcement learning to show that reputation mechanisms generate two coordination problems: agents need to learn how to coordinate on the meaning of existing reputations and collectively agree on a social norm to assign reputations to others based on their behavior. These coordination problems exhibit multiple equilibria, some of which effectively establish cooperation. When we train agents with a standard Q-learning algorithm in an environment with the presence of reputation mechanisms, convergence to undesirable equilibria is widespread. We propose two mechanisms to alleviate this: (i) seeding a proportion of the system with fixed agents that steer others towards good equilibria; and (ii), intrinsic rewards based on the idea of introspection, i.e., augmenting agents' rewards by an amount proportionate to the performance of their own strategy against themselves. A combination of these simple mechanisms is successful in stabilizing cooperation, even in a fully decentralized version of the problem where agents learn to use and assign reputations simultaneously. We show how our results relate to the literature in Evolutionary Game Theory, and discuss implications for artificial, human and hybrid systems, where reputations can be used as a way to establish trust and cooperation.

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。

↑