Multiphase sampling using expected value of information

S. de Bruin, D.E. Ballari, A.K. Bregt

Research output: Chapter in Book/Report/Conference proceedingConference paper

Abstract

This paper explores multiphase or infill sampling to reduce uncertainty after an initial sample has been taken and analysed to produce a map of the probability of some hazard. New observations are iteratively added by maximising the global expected value of information of the points. This is equivalent to minimisation of global misclassification costs. The method accounts for measurement error and different costs of type I and type II errors. Constraints imposed by a mobile sensor web can be accommodated using cost distances rather than Euclidean distances to decide which sensor moves to the next sample location. Calculations become demanding when multiple sensors move simultaneously. In that case, a genetic algorithm can be used to find sets of suitable new measurement locations. The method was implemented using R software for statistical computing and contributed libraries and it is demonstrated using a synthetic data set.
Original languageEnglish
Title of host publicationProceedings of the 7th International Symposium on Spatial Data Quality, Raising awareness of Spatial Data Quality, Coimbra, Portugal, 12 - 14 October, 2011
EditorsC.C. Fonte, L.M.S. Gonçalves, G. Gonçalves
Place of PublicationCoimbra, Portugal
PublisherUniversity of Coimbra
Pages65-70
Publication statusPublished - 2011
EventProceedings of the 7th International Symposium on Spatial Data Quality, Raising awareness of Spatial Data Quality, Coimbra, Portuga -
Duration: 12 Oct 201114 Oct 2011

Conference

ConferenceProceedings of the 7th International Symposium on Spatial Data Quality, Raising awareness of Spatial Data Quality, Coimbra, Portuga
Period12/10/1114/10/11

Fingerprint Dive into the research topics of 'Multiphase sampling using expected value of information'. Together they form a unique fingerprint.

  • Cite this

    de Bruin, S., Ballari, D. E., & Bregt, A. K. (2011). Multiphase sampling using expected value of information. In C. C. Fonte, L. M. S. Gonçalves, & G. Gonçalves (Eds.), Proceedings of the 7th International Symposium on Spatial Data Quality, Raising awareness of Spatial Data Quality, Coimbra, Portugal, 12 - 14 October, 2011 (pp. 65-70). Coimbra, Portugal: University of Coimbra.