Corpus ID: 203592142

A Generalized Training Approach for Multiagent Learning

  title={A Generalized Training Approach for Multiagent Learning},
  author={P. Muller and Shayegan Omidshafiei and M. Rowland and K. Tuyls and Julien P{\'e}rolat and S. Liu and D. Hennes and Luke Marris and Marc Lanctot and Edward Hughes and Z. Wang and G. Lever and N. Heess and T. Graepel and R{\'e}mi Munos},
  • P. Muller, Shayegan Omidshafiei, +12 authors Rémi Munos
  • Published 2020
  • Computer Science
  • ArXiv
  • This paper investigates a population-based training regime based on game-theoretic principles called Policy-Spaced Response Oracles (PSRO). PSRO is general in the sense that it (1) encompasses well-known algorithms such as fictitious play and double oracle as special cases, and (2) in principle applies to general-sum, many-player games. Despite this, prior studies of PSRO have been focused on two-player zero-sum games, a regime wherein Nash equilibria are tractably computable. In moving from… CONTINUE READING
    13 Citations

    Figures, Tables, and Topics from this paper.

    Pipeline PSRO: A Scalable Approach for Finding Approximate Nash Equilibria in Large Games
    • 1
    • PDF
    An Overview of Multi-Agent Reinforcement Learning from Game Theoretical Perspective
    Algorithms in Multi-Agent Systems: A Holistic Perspective from Reinforcement Learning and Game Theory
    Converging to Team-Maxmin Equilibria in Zero-Sum Multiplayer Games
    Double Oracle Algorithm for Computing Equilibria in Continuous Games
    αα-Rank: Practically Scaling α-Rank through Stochastic Optimisation
    • 2
    • Highly Influenced
    • PDF
    Navigating the Landscape of Games
    • 1


    Multiagent Evaluation under Incomplete Information
    • 15
    • PDF
    A Unified Game-Theoretic Approach to Multiagent Reinforcement Learning
    • 217
    • PDF
    Independent reinforcement learners in cooperative Markov games: a survey regarding coordination problems
    • 162
    • Highly Influential
    • PDF
    α-Rank: Multi-Agent Evaluation by Evolution
    • 29
    • PDF
    OpenSpiel: A Framework for Reinforcement Learning in Games
    • 24
    • PDF
    Choosing Samples to Compute Heuristic-Strategy Nash Equilibrium
    • 62
    • PDF
    A Generalised Method for Empirical Game Theoretic Analysis
    • 34
    • PDF
    Open-ended Learning in Symmetric Zero-sum Games
    • 37
    • PDF
    Evolutionary Reinforcement Learning for Sample-Efficient Multiagent Coordination
    • 6
    • PDF
    What evolutionary game theory tells us about multiagent learning
    • 93
    • PDF