Dynamics of Softmax Q-Learning in Two-Player Two-Action Games
Ardeshir KianercyAram GalstyanPublished in: CoRR (2011)
Keyphrases
- temporal difference learning
- game playing
- action selection
- nash equilibria
- perfect information
- reinforcement learning algorithms
- learning agents
- game theoretic
- stochastic games
- state action
- nash equilibrium
- reinforcement learning
- imperfect information
- minority game
- multi agent reinforcement learning
- multi player
- multiagent learning
- repeated games
- game theory
- single agent
- normal form games
- extensive form games
- multi agent
- state space
- learning agent
- cooperative
- model free
- incomplete information
- decision problems
- function approximation
- dynamical systems
- learning algorithm
- optimal policy
- video games
- educational games
- temporal difference
- multi agent systems
- evolutionary game theory
- computer games
- monte carlo
- reinforcement learning methods
- stochastic approximation
- action space
- dynamic environments
- game tree
- agent learns
- game play