Time Horizon Generalization in Reinforcement Learning: Generalizing Multiple Q-Tables in Q-Learning Agents.
Yasuyo HatchoKiyohiko HattoriKeiki TakadamaPublished in: J. Adv. Comput. Intell. Intell. Informatics (2009)
Keyphrases
- reinforcement learning
- multi agent
- action selection
- multi agent reinforcement learning
- function approximation
- single agent
- state action space
- learning agent
- cooperative
- multi agent systems
- reinforcement learning algorithms
- state space
- multiagent systems
- multiple agents
- learning agents
- learning algorithm
- database
- multiagent learning
- autonomous agents
- agent receives
- dynamic programming
- intelligent agents
- optimal policy
- agent learns
- robocup soccer
- temporal difference learning
- machine learning
- relational reinforcement learning
- eligibility traces
- model free
- mobile agents
- markov decision processes
- resource allocation
- databases
- multi agent environments
- continuous state and action spaces
- continuous state
- partial observability
- stochastic approximation
- stochastic games
- action space
- decision making
- policy iteration
- learning process
- dynamic environments
- learning classifier systems
- learning problems
- learning tasks