Robot Planning in Partially Observable Continuous Domains.

Josep M. Porta Matthijs T. J. Spaan Nikos A. Vlassis

Published in: BNAIC (2005)

Keyphrases

partially observable
continuous domains
reinforcement learning
state space
decision problems
partial observability
markov decision processes
dynamical systems
mobile robot

belief space
markov decision problems
belief state
partially observable markov decision processes
evolutionary computation
planning domains
infinite horizon
multi robot
reward function

evolution strategy
motion planning
learning algorithm
genetic algorithm
higher dimensional
machine learning
conditional independence
goal directed

computational intelligence
probability distribution
dynamic programming
search algorithm
bayesian networks
random walk
objective function
probabilistic model