On Improving Deep Reinforcement Learning for POMDPs.
Pengfei ZhuXin LiPascal PoupartPublished in: CoRR (2017)
Keyphrases
- reinforcement learning
- partially observable markov decision processes
- function approximation
- state space
- markov decision processes
- continuous state
- optimal policy
- multi agent
- partially observable
- policy search
- reinforcement learning algorithms
- model free
- temporal difference
- dynamic programming
- learning process
- supervised learning
- action selection
- learning algorithm
- robotic control
- reinforcement learning methods
- function approximators
- policy iteration algorithm
- learning problems
- control problems
- learning classifier systems
- finite state
- markov decision problems
- transfer learning
- search algorithm
- machine learning