Generalising Experience in Reinforcement Learning: Performance in Partially Observable Processes.
Carlos H. C. RibeiroPublished in: ICANNGA (1997)
Keyphrases
- partially observable
- reinforcement learning
- state space
- markov decision processes
- partial observability
- decision problems
- partially observable domains
- dynamical systems
- partially observable environments
- markov decision problems
- infinite horizon
- hidden state
- optimal policy
- function approximation
- partial observations
- reward function
- action models
- belief state
- temporal difference
- reinforcement learning algorithms
- markov decision process
- belief space
- model free
- dynamic programming
- decision making
- learning algorithm
- machine learning
- markov chain
- continuous state
- fully observable
- multi agent