Recurrent Macro Actions Generator for POMDP Planning.
Yuanchu LiangHanna KurniawatiPublished in: IROS (2023)
Keyphrases
- macro actions
- reinforcement learning
- markov decision processes
- state space
- markov decision problems
- finite state
- partially observable
- partially observable markov decision processes
- dynamic programming
- optimal policy
- planning domains
- planning problems
- belief state
- partially observable markov decision process
- function approximation
- markov decision process
- neural network
- multi agent
- predictive model
- dynamical systems
- plan library
- temporally extended
- reward function
- decision theoretic planning
- decision processes
- policy iteration
- search algorithm