Standard
Adaptive sensor placement for continuous spaces. /
Grant, James; Boukouvalas, Alexis; Griffiths, Ryan-Rhys et al.
Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019. Proceedings of Machine Learning Research, 2019. (Proceedings of Machine Learning Research; Vol. 97).
Research output: Contribution in Book/Report/Proceedings - With ISBN/ISSN › Conference contribution/Paper › peer-review
Harvard
Grant, J, Boukouvalas, A, Griffiths, R-R
, Leslie, D, Vakili, S & Munoz de Cote, E 2019,
Adaptive sensor placement for continuous spaces. in
Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019. Proceedings of Machine Learning Research, vol. 97, Proceedings of Machine Learning Research. <
http://proceedings.mlr.press/v97/>
APA
Grant, J., Boukouvalas, A., Griffiths, R-R.
, Leslie, D., Vakili, S., & Munoz de Cote, E. (2019).
Adaptive sensor placement for continuous spaces. In
Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019 (Proceedings of Machine Learning Research; Vol. 97). Proceedings of Machine Learning Research.
http://proceedings.mlr.press/v97/
Vancouver
Grant J, Boukouvalas A, Griffiths R-R
, Leslie D, Vakili S, Munoz de Cote E.
Adaptive sensor placement for continuous spaces. In Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019. Proceedings of Machine Learning Research. 2019. (Proceedings of Machine Learning Research).
Author
Grant, James ; Boukouvalas, Alexis ; Griffiths, Ryan-Rhys et al. /
Adaptive sensor placement for continuous spaces. Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019. Proceedings of Machine Learning Research, 2019. (Proceedings of Machine Learning Research).
Bibtex
@inproceedings{c6d3cf4a0ddd4fc9a78284cb30cf2612,
title = "Adaptive sensor placement for continuous spaces",
abstract = "We consider the problem of adaptively placing sensors along an interval to detect stochasticallygenerated events. We present a new formulation of the problem as a continuum-armed bandit problem with feedback in the form of partial observations of realisations of an inhomogeneous Poisson process. We design a solution method by combining Thompson sampling with nonparametric inference via increasingly granular Bayesian histograms and derive an O˜(T2/3) bound on the Bayesian regret in T rounds. This is coupled with the design of an efficent optimisation approach toselect actions in polynomial time. In simulations we demonstrate our approach to have substantially lower and less variable regret than competitor algorithms.",
author = "James Grant and Alexis Boukouvalas and Ryan-Rhys Griffiths and David Leslie and Sattar Vakili and {Munoz de Cote}, Enrique",
year = "2019",
month = jul,
day = "12",
language = "English",
series = "Proceedings of Machine Learning Research",
publisher = "Proceedings of Machine Learning Research",
booktitle = "Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019",
}
RIS
TY - GEN
T1 - Adaptive sensor placement for continuous spaces
AU - Grant, James
AU - Boukouvalas, Alexis
AU - Griffiths, Ryan-Rhys
AU - Leslie, David
AU - Vakili, Sattar
AU - Munoz de Cote, Enrique
PY - 2019/7/12
Y1 - 2019/7/12
N2 - We consider the problem of adaptively placing sensors along an interval to detect stochasticallygenerated events. We present a new formulation of the problem as a continuum-armed bandit problem with feedback in the form of partial observations of realisations of an inhomogeneous Poisson process. We design a solution method by combining Thompson sampling with nonparametric inference via increasingly granular Bayesian histograms and derive an O˜(T2/3) bound on the Bayesian regret in T rounds. This is coupled with the design of an efficent optimisation approach toselect actions in polynomial time. In simulations we demonstrate our approach to have substantially lower and less variable regret than competitor algorithms.
AB - We consider the problem of adaptively placing sensors along an interval to detect stochasticallygenerated events. We present a new formulation of the problem as a continuum-armed bandit problem with feedback in the form of partial observations of realisations of an inhomogeneous Poisson process. We design a solution method by combining Thompson sampling with nonparametric inference via increasingly granular Bayesian histograms and derive an O˜(T2/3) bound on the Bayesian regret in T rounds. This is coupled with the design of an efficent optimisation approach toselect actions in polynomial time. In simulations we demonstrate our approach to have substantially lower and less variable regret than competitor algorithms.
M3 - Conference contribution/Paper
T3 - Proceedings of Machine Learning Research
BT - Proceedings of the 36 th International Conference on Machine Learning, Long Beach, California, PMLR 97, 2019
PB - Proceedings of Machine Learning Research
ER -
