Dynamic deep-reinforcement-learning algorithm in Partially Observed Markov Decision Processes.
Saki OmiHyo-Sang ShinNamhoon ChoAntonios TsourdosPublished in: CoRR (2023)
Keyphrases
- markov decision processes
- partially observed
- state space
- optimal policy
- dynamic programming
- finite state
- reinforcement learning
- policy iteration
- decision theoretic planning
- expected reward
- decision processes
- action space
- average cost
- transition matrices
- reachability analysis
- infinite horizon
- reinforcement learning algorithms
- action sets
- average reward
- risk sensitive
- factored mdps
- partially observable
- planning under uncertainty
- data mining
- finite horizon
- model based reinforcement learning
- decision problems
- dynamical systems
- state and action spaces
- dynamic environments
- total reward
- semi markov decision processes