Flow-based Recurrent Belief State Learning for POMDPs.

Xiaoyu Chen Yao Mark Mu Ping Luo Shengbo Li Jianyu Chen

Published in: ICML (2022)

Keyphrases

belief state
partially observable
state space
reinforcement learning
learning algorithm
partially observable markov decision processes
continuous state
prior knowledge
active learning
model selection
dynamical systems
least squares
markov chain
finite state
dynamic bayesian networks
partial observability