Fictitious play in networks

Ewerhart, Christian; Valkanova, Kremena

doi:10.1016/j.geb.2020.06.006

Cited by 9 publications

(6 citation statements)

References 80 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Additionally, every potential game has the fictitious play property and converges to equilibrium [25]. This means that there will be two equilibrium points according to the beliefs of the players.…”

Section: Fig 3: Lte Selection Example Of the Gamementioning

confidence: 99%

DSRC or LTE? Selecting the Best Medium for V2I Communication Using Game Theory

Spyrou

Anagnostopoulou

Stylios

2021

IFIP Advances in Information and Communication Technology

View full text Add to dashboard Cite

Vehicular communication is a very challenging and essential research area capable of supporting safety and routing decision-making. Vehicle to Infrastructure (V2I) communication often refers to communication between vehicles and Road Side Units (RSU), and recently several technologies have been developed to support it, such as ZigBee, Wi-Fi, GSM, Long Term Evolution (LTE), and 802.11p Direct Short Range Communication (DSRC). In this field, there is a competition between wireless DSRC and cellular LTE to define the most efficient type of communication. This paper aims to analyze the strengths and weaknesses of the DSRC and LTE to evaluate their performances and select the right technology for communication between vehicles and RSUs. Therefore, a vehicle equipped with both LTE and DSRC modules is assumed, and we propose a game-theoretic formulation to select the most efficient type of communication. The proposed formulation results in two equilibria; based on them, the vehicle and the RSU select the same communication module. Here it presents the correlated equilibrium when a trusted source makes the decision, and it discusses the two equilibria as a potential game formulation.

show abstract

Section: Fig 3: Lte Selection Example Of the Gamementioning

confidence: 99%

DSRC or LTE? Selecting the Best Medium for V2I Communication Using Game Theory

Spyrou

Anagnostopoulou

Stylios

2021

IFIP Advances in Information and Communication Technology

View full text Add to dashboard Cite

show abstract

“…SFP and its variants have recently attracted a lot of attention in AI research [36,37,42,18,17]. There is a significant literature that analyze SFP in different models [3,7,21,19], and the paper that is most closely related to our work is [12]. Fudenberg and Takahashi [12] also examines the heterogeneity issue and anticipate belief homogenization in the limit under 2-population settings.…”

Section: Introductionmentioning

confidence: 99%

“…On the other hand, networked multi-agent learning constitutes one of the current frontiers in AI and ML research [43,30,16]. Recent theoretical advances on network games provide conditions for learning behaviors to be not chaotic [6,34], and investigate the convergence of Q-learning and continuous-time FP in the case of network competitions [7,28]. However, [7,28] consider that there is only one agent on each vertex, and hence their models are essentially for homogeneous systems.…”

Section: Introductionmentioning

confidence: 99%

Heterogeneous Beliefs and Multi-Population Learning in Network Games

Hu¹,

Soh²,

Piliouras³

2023

Preprint

View full text Add to dashboard Cite

The effect of population heterogeneity in multi-agent learning is practically relevant but remains far from being well-understood. Motivated by this, we introduce a model of multi-population learning that allows for heterogeneous beliefs within each population and where agents respond to their beliefs via smooth fictitious play (SFP). We show that the system state -a probability distribution over beliefs -evolves according to a system of partial differential equations. We establish the convergence of SFP to Quantal Response Equilibria in different classes of games capturing both network competition as well as network coordination. We also prove that the beliefs will eventually homogenize in all network games. Although the initial belief heterogeneity disappears in the limit, we show that it plays a crucial role for equilibrium selection in the case of coordination games as it helps select highly desirable equilibria. Contrary, in the case of network competition, the resulting limit behavior is independent of the initialization of beliefs, even when the underlying game has many distinct Nash equilibria.

show abstract

“…The study of Multi-Agent Learning in competitive settings has achieved a number of successes within the context of zero-sum games and their network variants These games model perfect competition between agents, yielding an underlying structure which make them amenable for studying multi-agent learning. In particular it is known that certain learning dynamics asymptotically converge to an equilibrium in network zero-sum games (Ewerhart and Valkanova 2020;Leonardos, Piliouras, and Spendlove 2021; Copyright © 2024, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.…”

Section: Introductionmentioning

confidence: 99%

Stability of Multi-Agent Learning in Competitive Networks: Delaying the Onset of Chaos

Hussain,

Belardinelli

2024

AAAI

View full text Add to dashboard Cite

The behaviour of multi agent learning in competitive network games is often studied within the context of zero sum games, in which convergence guarantees may be obtained. However, outside of this class the behaviour of learning is known to display complex behaviours and convergence cannot be always guaranteed. Nonetheless, in order to develop a complete picture of the behaviour of multi agent learning in competitive settings, the zero sum assumption must be lifted. Motivated by this we study the Q Learning dynamics, a popular model of exploration and exploitation in multi agent learning, in competitive network games. We determine how the degree of competition, exploration rate and network connectivity impact the convergence of Q Learning. To study generic competitive games, we parameterise network games in terms of correlations between agent payoffs and study the average behaviour of the Q Learning dynamics across all games drawn from a choice of this parameter. This statistical approach establishes choices of parameters for which Q Learning dynamics converge to a stable fixed point. Differently to previous works, we find that the stability of Q Learning is explicitly dependent only on the network connectivity rather than the total number of agents. Our experiments validate these findings and show that, under certain network structures, the total number of agents can be increased without increasing the likelihood of unstable or chaotic behaviours.

show abstract

Fictitious play in networks

Cited by 9 publications

References 80 publications

DSRC or LTE? Selecting the Best Medium for V2I Communication Using Game Theory

DSRC or LTE? Selecting the Best Medium for V2I Communication Using Game Theory

Heterogeneous Beliefs and Multi-Population Learning in Network Games

Stability of Multi-Agent Learning in Competitive Networks: Delaying the Onset of Chaos

Contact Info

Product

Resources

About