Partially Observable Hierarchical Reinforcement Learning with AI Planning (Student Abstract).
Brandon RozekJunkyu LeeHarsha KokelMichael KatzShirin SohrabiPublished in: AAAI (2024)
Keyphrases
- partially observable
- ai planning
- hierarchical reinforcement learning
- reinforcement learning
- action models
- planning domains
- reward function
- state space
- planning problems
- heuristic search
- plan existence
- markov decision processes
- planning under uncertainty
- model free
- orders of magnitude
- web service composition
- markov decision process
- planning systems
- integer programming
- decision problems
- markov decision problems
- infinite horizon
- dynamical systems
- domain independent
- function approximation
- service composition
- search strategies
- domain specific
- dynamic programming
- average reward
- reinforcement learning algorithms
- learning algorithm
- optimal policy
- model checking
- belief state
- constraint satisfaction
- machine learning
- multiple agents
- search space
- finite state
- multi agent