Perception-Aware Point-Based Value Iteration for Partially Observable Markov Decision Processes.

Mahsa Ghasemi Ufuk Topcu

Published in: IJCAI (2019)

Keyphrases

point based value iteration
partially observable markov decision processes
belief state
belief space
reinforcement learning
finite state
continuous state
dynamical systems
dynamic programming
state space
optimal policy
decision problems
planning under uncertainty
domain specific
sufficient conditions