TY - GEN
T1 - Efficient deterministic approximation algorithms for non-myopic value of information in graphical models
AU - Radovilsky, Yan
AU - Shattah, Guy
AU - Shimony, Solomon Eyal
PY - 2006/1/1
Y1 - 2006/1/1
N2 - Agents operating in the real world need to handle both uncertainty and resource constraints. Typical problems in this domain are optimization of sequences of observations, and optimal allocation of computation tasks during reasoning and search (also known as meta-reasoning). In both domains, a crucial issue is value of information, a quantity hard to compute in general, and thus usually estimated using severe assumptions, such as myopic and independence of information sources. This paper extends recent work on non-myopic value of information in graphical models, that assumed a chain-shaped graph and exact measurements. Suitably relaxing the assumption of exact measurements still allows for a provably close approximation of the optimal subset (of observations) selection, and for approximating the optimal conditional plan. The method is shown to be efficient and to provide a significant advantage in expected reward over the myopic and greedy value of information scheme.
AB - Agents operating in the real world need to handle both uncertainty and resource constraints. Typical problems in this domain are optimization of sequences of observations, and optimal allocation of computation tasks during reasoning and search (also known as meta-reasoning). In both domains, a crucial issue is value of information, a quantity hard to compute in general, and thus usually estimated using severe assumptions, such as myopic and independence of information sources. This paper extends recent work on non-myopic value of information in graphical models, that assumed a chain-shaped graph and exact measurements. Suitably relaxing the assumption of exact measurements still allows for a provably close approximation of the optimal subset (of observations) selection, and for approximating the optimal conditional plan. The method is shown to be efficient and to provide a significant advantage in expected reward over the myopic and greedy value of information scheme.
UR - http://www.scopus.com/inward/record.url?scp=34548128871&partnerID=8YFLogxK
U2 - 10.1109/ICSMC.2006.385249
DO - 10.1109/ICSMC.2006.385249
M3 - Conference contribution
AN - SCOPUS:34548128871
SN - 1424401003
SN - 9781424401000
T3 - Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics
SP - 2559
EP - 2564
BT - 2006 IEEE International Conference on Systems, Man and Cybernetics
PB - Institute of Electrical and Electronics Engineers
T2 - 2006 IEEE International Conference on Systems, Man and Cybernetics
Y2 - 8 October 2006 through 11 October 2006
ER -