POMDPPartially Observable Markov Decision Process (artificial intelligence)
References in periodicals archive ?
6] first proposed the looping prediction suffix trees (LPST) in the deterministic POMDP environment, which can map the long-term dependent histories onto a finite LPST.
Section 4 models general autonomous driving decision-making process in a POMDP, while the approximations and the simplified solver are described in Section 5.
Cross Product MDP: POMDP with <S,A,T,R,O,Z> and policy graph with the node set N cross-product MDP <[bar.
However, despite the work of Papadimitriou and Tsitsiklis [1987] and a few others, there were many variants of POMDP problems for which it had not been proved that finding tractable exact solutions or provably good approximations is hard.
Dissanayake, "Wheelchair driver assistance and intention prediction using POMDPs," in Proceedings of the International Conference on Intelligent Sensors, Sensor Networks and Information Processing (ISSNIP '07), pp.
A Dec-POMDP is a generalization to multiple agents of the well-known POMDP model for single-agent planning under uncertainty (19).
10]) actions) spaces are beyond the reach of current state-of-the-art POMDP solvers and algorithms.
Participation in IPPC-2014 offered a rematch of the top two competitors from the respective MDP and POMDP tracks of IPPC-2011 using variants of Monte Carlo tree search (MCTS) and online value iteration.
As mentioned in [6], it is well known that for each POMDP, all information that is useful for making decisions can be encapsulated in the belief vector.
Performing POMDP planning in such domains is generally computationally challenging, since a curriculum may consist of many skills, each of which may be known or not known, and the skills are not independent.
In order to study the above system, it is modeled as a POMDP since its state cannot be completely observed.
The first framework that I consider is the decentralized POMDP (Dec-POMDP) (Bernstein et al.