Online Reinforcement Learning of Optimal Threshold Policies for Markov Decision Processes.
Arghyadip RoyVivek S. BorkarAbhay KarandikarPrasanna ChaporkarPublished in: IEEE Trans. Autom. Control. (2022)
Keyphrases
- markov decision processes
- optimal policy
- reinforcement learning
- average cost
- total reward
- policy iteration algorithm
- stationary policies
- dynamic programming
- discounted reward
- markov decision process
- average reward
- finite horizon
- reinforcement learning algorithms
- finite state
- reward function
- state space
- control policies
- policy iteration
- expected reward
- action sets
- discount factor
- infinite horizon
- decision problems
- control policy
- decentralized control
- model based reinforcement learning
- action space
- state and action spaces
- decision theoretic planning
- planning under uncertainty
- transition matrices
- decision processes
- hierarchical reinforcement learning
- optimality criterion
- approximate dynamic programming
- policy evaluation
- factored mdps
- markov decision problems
- long run
- partially observable
- markov games
- macro actions
- function approximation
- optimal control
- stochastic games
- temporal difference
- model free
- continuous state spaces
- actor critic
- continuous state
- sufficient conditions
- lost sales
- multi agent
- partially observable markov decision processes
- optimal solution
- action selection
- machine learning