Select Lab Publications


Near-optimal Value of Information in Graphical Models (2005)

By: Andreas Krause and Carlos Guestrin

Abstract: A fundamental issue in real-world systems, such as sensor networks, is the selection of observations which most effectively reduce uncertainty. More specifically, we address the long standing problem of nonmyopically selecting the most informative subset of variables in a graphical model. We present the first efficient randomized algorithm providing a constant factor (1-1/e-eps) approximation guarantee for any eps > 0 with high confidence. The algorithm leverages the theory of submodular functions, in combination with a polynomial bound on sample complexity. We furthermore prove that no polynomial time algorithm can provide a constant factor approximation better than (1 - 1/e) unless P = NP. Finally, we provide extensive evidence of the effectiveness of our method on two complex real-world datasets.

Download Information
Andreas Krause and Carlos Guestrin (2005). "Near-optimal Value of Information in Graphical Models." Conference on Uncertainty in Artificial Intelligence (UAI). Winner of the Best Paper Runner-up Award. pdf   talk        
BibTeX citation

@inproceedings{Krause+Guestrin:uai05infogain,
author = {Andreas Krause and Carlos Guestrin},
title = {Near-optimal Value of Information in Graphical Models},
booktitle = {Conference on Uncertainty in Artificial Intelligence (UAI)},
month = {July},
year = {2005},
wwwfilebase = {uai2005-krause-guestrin},
wwwtopic = {Observation Selection},
wwwaward = {Winner of the Best Paper Runner-up Award}
}



full list