Sciweavers

13 search results - page 1 / 3
» Point-Based Value Iteration for Continuous POMDPs
Sort
View
JMLR
2006
116views more  JMLR 2006»
13 years 11 months ago
Point-Based Value Iteration for Continuous POMDPs
We propose a novel approach to optimize Partially Observable Markov Decisions Processes (POMDPs) defined on continuous spaces. To date, most algorithms for model-based POMDPs are ...
Josep M. Porta, Nikos A. Vlassis, Matthijs T. J. S...
AAAI
2008
14 years 1 months ago
Generalized Point Based Value Iteration for Interactive POMDPs
We develop a point based method for solving finitely nested interactive POMDPs approximately. Analogously to point based value iteration (PBVI) in POMDPs, we maintain a set of bel...
Prashant Doshi, Dennis Perez

Publication
273views
13 years 6 months ago
Monte Carlo Value Iteration for Continuous-State POMDPs
Partially observable Markov decision processes (POMDPs) have been successfully applied to various robot motion planning tasks under uncertainty. However, most existing POMDP algo...
Haoyu Bai, David Hsu, Wee Sun Lee, and Vien A. Ngo
AI
2006
Springer
14 years 2 months ago
Belief Selection in Point-Based Planning Algorithms for POMDPs
Abstract. Current point-based planning algorithms for solving partially observable Markov decision processes (POMDPs) have demonstrated that a good approximation of the value funct...
Masoumeh T. Izadi, Doina Precup, Danielle Azar
IJCAI
2007
14 years 12 days ago
Improving Anytime Point-Based Value Iteration Using Principled Point Selections
Planning in partially-observable dynamical systems (such as POMDPs and PSRs) is a computationally challenging task. Popular approximation techniques that have proved successful ar...
Michael R. James, Michael E. Samples, Dmitri A. Do...