Decentralized Q-Learning with Constant Aspirations in Stochastic Games.
Bora YongacogluGürdal ArslanSerdar YükselPublished in: ACSSC (2019)
Keyphrases
- stochastic games
- multi agent
- multiagent reinforcement learning
- reinforcement learning algorithms
- multi agent reinforcement learning
- state action
- single agent
- cooperative
- reinforcement learning
- learning agent
- markov decision processes
- state space
- nash equilibria
- model free
- rl algorithms
- average reward
- optimal policy
- learning automata
- autonomous agents
- repeated games
- multiple agents
- function approximation
- least squares
- path finding
- imperfect information
- learning capabilities
- dynamic environments
- genetic algorithm
- action selection
- search space