Monte Carlo POMDP

In the class of Markov decision process algorithms, the Monte Carlo POMDP (MC-POMDP) is the particle filter version for the partially observable Markov decision process (POMDP) algorithm. In MC-POMDP, particles filters are used to update and approximate the beliefs, and the algorithm is applicable to continuous valued states, actions, and measurements.[1]

References

  1. Thrun, S.; Burgard, W.; Fox, D. (2005). Probabilistic Robotics. Cambridge: The MIT Press. ISBN 0-262-20162-3.
This article is issued from Wikipedia - version of the 7/22/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.