Skip to main content
placeholder image

Spatial data compression via adaptive dispersion clustering

Journal Article


Abstract


  • Spatial Dispersion Clustering (ASDC), a new method of spatial data compression, is specifically designed to reduce the size of a spatial dataset in order to facilitate subsequent spatial prediction. Unlike traditional data and image compression methods, the goal of ASDC is to create a new dataset that will be used as input into spatial-prediction methods, such as traditional kriging or Fixed Rank Kriging, where using the full dataset may be computationally infeasible. ASDC can be classified as a lossy compression method and is based on spectral clustering. It aims to produce contiguous spatial clusters and to preserve the spatial-correlation structure of the data so that the loss of predictive information is minimal. An extensive simulation study demonstrates the predictive performance of these adaptively compressed datasets for several scenarios. ASDC is compared to two other data-reduction schemes, one using local neighborhoods and one using simple binning. An application to remotely sensed sea-surface-temperature data is also presented, and computational costs are discussed.

Authors


  •   Marchetti, Yuliya (external author)
  •   Nguyen, Hai (external author)
  •   Braverman, Amy (external author)
  •   Cressie, Noel A.

Publication Date


  • 2018

Citation


  • Marchetti, Y., Nguyen, H., Braverman, A. & Cressie, N. (2018). Spatial data compression via adaptive dispersion clustering. Computational Statistics and Data Analysis, 117 138-153.

Scopus Eid


  • 2-s2.0-85028854928

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers1/1992

Number Of Pages


  • 15

Start Page


  • 138

End Page


  • 153

Volume


  • 117

Place Of Publication


  • Netherlands

Abstract


  • Spatial Dispersion Clustering (ASDC), a new method of spatial data compression, is specifically designed to reduce the size of a spatial dataset in order to facilitate subsequent spatial prediction. Unlike traditional data and image compression methods, the goal of ASDC is to create a new dataset that will be used as input into spatial-prediction methods, such as traditional kriging or Fixed Rank Kriging, where using the full dataset may be computationally infeasible. ASDC can be classified as a lossy compression method and is based on spectral clustering. It aims to produce contiguous spatial clusters and to preserve the spatial-correlation structure of the data so that the loss of predictive information is minimal. An extensive simulation study demonstrates the predictive performance of these adaptively compressed datasets for several scenarios. ASDC is compared to two other data-reduction schemes, one using local neighborhoods and one using simple binning. An application to remotely sensed sea-surface-temperature data is also presented, and computational costs are discussed.

Authors


  •   Marchetti, Yuliya (external author)
  •   Nguyen, Hai (external author)
  •   Braverman, Amy (external author)
  •   Cressie, Noel A.

Publication Date


  • 2018

Citation


  • Marchetti, Y., Nguyen, H., Braverman, A. & Cressie, N. (2018). Spatial data compression via adaptive dispersion clustering. Computational Statistics and Data Analysis, 117 138-153.

Scopus Eid


  • 2-s2.0-85028854928

Ro Metadata Url


  • http://ro.uow.edu.au/eispapers1/1992

Number Of Pages


  • 15

Start Page


  • 138

End Page


  • 153

Volume


  • 117

Place Of Publication


  • Netherlands