Estimating Disentangled Belief about Hidden State and Hidden Task for Meta-RL.

Kei Akuzawa Yusuke Iwasawa Yutaka Matsuo

Published in: CoRR (2021)

Keyphrases

hidden state
reinforcement learning
hidden markov models
markov models
partially observable
dynamical systems

partially observable markov decision processes
belief state
belief space
state space
markov decision processes
learning algorithm

model free
fully observable
optimal control
dynamic programming
markov model
belief revision

optimal policy
orders of magnitude
action space
multi agent
information retrieval