Point-Based Monte Carto Online Planning in POMDPs

• DOI: 10.4028/www.scientific.net/amr.846-847.1388
• Publication Date: 2013-11-21T15:57:21Z
• Authors: Bo Wu; Yan Peng Feng; Hong Yan Zheng
• Keywords: Simplex

The online planning and learning in partially observable Markov decision processes are often intractable because belief states space has two curses: dimensionality history. In order to address this problem, paper proposes a point-based Monte Carto approach POMDPs. This involves performing value backup at specific reachable points, rather than over the entire simplex, speed up computation processes. Then Carlo tree search algorithm is exploited share of actions across each subtree so as minimise mean squared error. experimental results show that proposed effective real-time system.