Evolution of boundedly rational learning in games
People make strategic decisions multiple times a day. We act strategically in negotiations, when we coordinate our actions with others, or when we choose with whom to cooperate. The resulting dynamics can be studied with evolutionary game theory. This framework explores how people adapt their decisions over time, in light of how effective their strategies have proven to be. A crucial quantity in respective models is the strength of selection. This quantity regulates how likely individuals switch to a better strategy when given the choice. The larger the selection strength, the more biased is the learning process in favor of strategies with large payoffs. Therefore, this quantity is often interpreted as a measure of rationality. Traditionally, most models take selection strength to be a fixed parameter. Instead, here we allow the individuals' strategies and their selection strength to co-evolve. The endpoints of this co-evolutionary process depend on the strategic interaction in place. In many prisoner's dilemmas, selection strength increases indefinitely, as one may expect. However, in snowdrift or stag-hunt games, it can either converge to a finite value, or we observe evolutionary branching altogether - such that different individuals arrive at different selection strengths. Overall, this work sheds light on how evolution might shape learning mechanisms for social behavior. It suggests that boundedly rational learning is not only a by-product of cognitive constraints. Instead it might also evolve as a means to gain strategic advantages.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Learning to Play General-Sum Games Against Multiple Boundedly Rational Agents
We study the problem of training a principal in a multi-agent general-sum game using reinforcement learning (RL). Learning a robust principal policy requires anticipating the worst possible strategic responses of other a…
Decision MakingMulti-agent Reinforcement LearningReinforcement Learning (RL)Cournot duopoly games with isoelastic demands and diseconomies of scale
In this discussion draft, we investigate five different models of duopoly games, where the market is assumed to have an isoelastic demand function. Moreover, quadratic cost functions reflecting decreasing returns to scal…
Machine Learning Techniques for Stackelberg Security Games: a Survey
The present survey aims at presenting the current machine learning techniques employed in security games domains. Specifically, we focused on papers and works developed by the Teamcore of University of Southern Californi…
BIG-bench Machine LearningSurveyLarge Scale Learning of Agent Rationality in Two-Player Zero-Sum Games
With the recent advances in solving large, zero-sum extensive form games, there is a growing interest in the inverse problem of inferring underlying game parameters given only access to agent actions. Although a recent w…
Soft-Bellman Equilibrium in Affine Markov Games: Forward Solutions and Inverse Learning
Markov games model interactions among multiple players in a stochastic, dynamic environment. Each player in a Markov game maximizes its expected total discounted reward, which depends upon the policies of the other playe…
OpenAI Gym