Memory-based Deep Reinforcement Learning for POMDP.
Lingheng MengRob GorbetDana KulicPublished in: CoRR (2021)
Keyphrases
- reinforcement learning
- continuous state
- hidden state
- partially observable
- function approximation
- optimal policy
- markov decision processes
- partially observable markov decision processes
- state space
- model free reinforcement learning
- temporal difference
- learning algorithm
- policy evaluation
- partially observable markov decision process
- model free
- reinforcement learning algorithms
- markov decision process
- action selection
- dynamic programming
- machine learning
- control problems
- supervised learning
- multi agent
- least squares
- policy iteration
- control policy
- reinforcement learning methods
- planning under uncertainty
- reward function
- computational complexity
- memory based learning
- sufficient conditions
- partially observable domains
- finite state