Globally Optimal Hierarchical Reinforcement Learning for Linearly-Solvable Markov Decision Processes.
Guillermo InfanteAnders JonssonVicenç GómezPublished in: AAAI (2022)
Keyphrases
- globally optimal
- markov decision processes
- hierarchical reinforcement learning
- state abstraction
- average reward
- reinforcement learning
- reward function
- policy iteration
- markov decision process
- finite state
- graph cuts
- optimal policy
- state space
- reinforcement learning algorithms
- model free
- np hard
- transition matrices
- function approximation
- dynamic programming
- infinite horizon
- partially observable
- discounted reward
- machine learning
- average cost
- action space
- decision theoretic planning
- computational complexity
- temporal difference
- multistage
- multi agent