Integration of fuzzy and geostatistical models for estimating missing multivariate observations

  • Authors:

  • Tuan D. Pham

  • Affiliations:

  • Bioinformatics Applications Research Center and School of Information Technology, James Cook University, Townsville, QLD, Australia

  • Venue:
  • FS'05 Proceedings of the 6th WSEAS international conference on Fuzzy systems
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

The estimation of missing observations is an important research field which has practical applications in many science and engineering disciplines. In analyzing the variability of a particular data set which can be spatially related, classical statistical methods make no use of this type of information; whereas geostatistics accomodates the spatial information of the data set in its regression analysis for estimating missing observations or unknown data. This paper incorporates the modeling of fuzzy protoptyes in the cokriging system of geostatistics in order to improve the accuracy of the estimates and alleviate the computational complexity of cokriging.