Generating Memoryless Policies Faster Using Automatic Temporal Abstractions for Reinforcement Learning with Hidden State.
Erkin ÇildenFaruk PolatPublished in: ICTAI (2013)
Keyphrases
- hidden state
- reinforcement learning
- temporal abstractions
- optimal policy
- partially observable markov decision processes
- hidden markov models
- markov models
- markov decision problems
- partially observable
- state space
- temporal pattern mining
- reward function
- learning algorithm
- markov decision processes
- multi agent
- dynamical systems
- machine learning
- dynamic programming
- decision problems
- statistical analysis
- computational complexity