If multi-agent learning is the answer, what is the question?

By Yoav Shoham, Rob Powers, Trond Grenager. 2007

470 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Management Science and Operations Research, Reinforcement Learning, Game & Decision Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Management Science and Operations Research

In Reinforcement Learning

In Game & Decision Theory

Citations in the corpus

Multi-agent Reinforcement Learning in Sequential Social Dilemmas (2017, background): “Most previous work on finding policies for Markov games takes the prescriptive view of multiagent learning [26]: that is, it attempts to answer “what should each agent do?” Several algorithms and analyses have been developed for the two-player zero-sum case [22, 27, 28, 29, 30].”
Empirically Evaluating Multiagent Learning Algorithms (2014, background): “ronment, the best policy to follow depends on the actions taken by the opponent, and thus on the ways in which the opponent’s future behavior will be affected by the learner’s 1 For alternatives, see Shoham et al. (2007)—who called the ap proach that we espouse the “prescriptive, noncooperative agen”
Learning to Communicate to Solve Riddles with Deep Distributed Recurrent Q-Networks (2016, methodology): “We ﬁrst evaluate DDRQN for n = 10 and compare it with tabular Q-learning.”
Multiagent Learning: Basics, Challenges, and Prospects (2012, methodology): “Articles that give a good overview of MAL methods and techniques developed during this second phase are, for example, Busoniu, Babuška, and De Schutter (2008); ’t Hoen et al. (2006), and Shoham, Powers, and Grenager (2007).”
A Unified Game-Theoretic Approach to Multiagent Reinforcement Learning (2017, background): “Inspired by recent success in Deep RL, we are now seeing a renewed interest in multiagent reinforcement learning (MARL) [90, 17, 99].”
A survey and critique of multiagent deep reinforcement learning (2018, methodology): “There is a long-standing history of combining game theory and MAL [43,233,289].”
Safe, Multi-Agent, Reinforcement Learning for Autonomous Driving (2016, background): “As noted in [30], learning in multi-agent setting is inherently more complex than in the single agent setting.”
Evolutionary Dynamics of Multi-Agent Learning: A Survey (2015, background): “They show how research on the evolutionary framework contributes to each of the five agendas of research identified by Shoham et al. (2007).”
Open-Ended Learning Leads to Generally Capable Agents (2021, neutral): “This complex interaction is central to the fields of Game Theory andmulti-agent RL (Balduzzi et al., 2019; Shoham et al., 2007).”
A multi-agent reinforcement learning model of common-pool resource appropriation (2017, unknown): “[34].”
Deep Reinforcement Learning for Multiagent Systems: A Review of Challenges, Solutions, and Applications (2018, background): “•Use the descriptive agenda [104] to describe the behaviors emerging when agents learn in the presence of other learning agents.”
What evolutionary game theory tells us about multiagent learning (2007, methodology): “In this section we will use EGT to illustrate our reaction to the analysis of multiagent learning presented in [20].”
Arena: A General Evaluation Platform and Building Toolkit for Multi-Agent Intelligence (2019, background): “Surveys of multi-agent intelligence research can be found here [13, 71, 84, 95].”
Deep Reinforcement Learning: An Overview (2017, background): “(2013) proposed to model the adaptive traffic signal control problem as a multiple player stochastic game, and solve it with the approach of multi-agent RL (Shoham et al., 2007; Busoniu et al., 2008).”
Social reward shaping in the prisoner's dilemma (2008, background): “In this work, we are concerned with what is called the “prescriptive non-cooperative” agenda [7].”
A Survey of Learning in Multiagent Environments: Dealing with Non-Stationarity (2017, background): “…systems such as: team vs concurrent learning (Panait and Luke, 2005); temporal difference, game theory and direct policy search (Busoniu et al., 2010); model-based, model-free and regret minimization (Shoham et al., 2007); and joint action, gradient, Nash and other learners (see Weiss, 2013, chap.”
Artificial Intelligence, Algorithmic Pricing, and Collusion (2020, unknown): “1 − δ 37 A similar problem emerges when δ is very close to 1.”
Theoretical considerations of potential-based reward shaping for multi-agent systems (2011, background): “Instead some compromise must be made, typically the system is designed aiming to converge to a Nash Equilibrium [24].”
Multi-agent Reinforcement Learning: An Overview (2010, methodology): “Shoham et al. [108] have critically examined the focus of MARL research on game-theoretic equilibria, using a selection of MARL algorithms to illustrate their arguments.”