Theoretically Guaranteed Policy Improvement Distilled from Model-Based Planning.
Chuming LiRuonan JiaJie LiuYinmin ZhangYazhe NiuYaodong YangYu LiuWanli OuyangPublished in: CoRR (2023)
Keyphrases
- action selection
- optimal policy
- planning problems
- heuristic search
- reinforcement learning problems
- blocks world
- ai planning
- case study
- genetic algorithm
- goal oriented
- information systems
- decision theoretic
- domain independent
- linear programming
- knowledge base
- multi agent
- data driven
- partially observable
- mixed initiative
- plan generation
- state space
- markov decision problems
- sequential decision making
- policy making
- significant improvement