Limiting Dynamics for Q-Learning with Memory One in Symmetric Two-Player, Two-Action Games.
Janusz M. MeylahnL. JanssenPublished in: Complex. (2022)
Keyphrases
- nash equilibria
- perfect information
- reinforcement learning algorithms
- stochastic games
- action selection
- learning agents
- state action
- nash equilibrium
- game theoretic
- reinforcement learning
- multi agent
- repeated games
- game theory
- imperfect information
- minority game
- game playing
- extensive form games
- cooperative
- multi agent reinforcement learning
- multiagent learning
- single agent
- action sets
- state space
- normal form games
- multi player
- incomplete information
- model free
- function approximation
- dynamic model
- memory requirements
- markov decision processes
- video games
- main memory
- learning algorithm
- dynamical systems
- temporal difference
- computer games
- influence diagrams
- learning rate
- game play
- dynamic programming
- multiple agents
- agent learns
- complex environments
- evaluation function
- online game
- evolutionary game theory
- potential field
- board game
- solution concepts
- stochastic approximation
- multiagent systems