Extended Q-Learning: Reinforcement Learning Using Self-Organized State Space.
Shuichi EnokidaTakeshi OhashiTakaichi YoshidaToshiaki EjimaPublished in: RoboCup (2000)
Keyphrases
- state space
- reinforcement learning
- reinforcement learning algorithms
- optimal policy
- function approximation
- markov decision processes
- continuous state spaces
- action space
- heuristic search
- dynamic programming
- markov decision process
- control problems
- state variables
- particle filter
- model free
- complex systems
- partially observable
- reward function
- markov chain
- learning process
- machine learning
- multi agent
- learning agent
- temporal difference learning
- planning problems
- temporal difference
- state abstraction
- hierarchical reinforcement learning
- eligibility traces
- learning algorithm
- mountain car
- macro actions
- state action
- function approximators
- stochastic approximation
- continuous state
- action selection
- state transition
- policy search
- learning classifier systems
- dynamical systems
- dynamic environments
- reward shaping
- cooperative