Abstract

Some specific random fields have been studied by many researchers whose finite-dimensional marginal distributions are multivariate closed skewnormal or multivariate extended skew-t, in time and spatial domains. In this paper, a necessary and sufficient condition is provided for applicability of such random field in spatial interpolation, based on the marginal distributions. Two deficiencies of the random fields generated by some well-known multivariate distributions are pointed out and in contrast, a suitable skew and heavy tailed random field is proposed. The efficiency of the proposed random field is illustrated through the interpolation of a real data.

Highlights

  • In recent years, random fields (RFs) have been successfully applied, as a statistical model, to the analysis of biological sequences, text and image processing, as well as many areas of computer vision and artificial intelligence

  • We show that the RFs generated by some versions of multivariate SN, Closed Skew-Normal (CSN) and Extended Skew-t (EST) distributions have two weakness in application to the spatial interpolation

  • Our findings show that definition of a RF based on multivariate CSN and EST distributions as studied by the corresponding researchers, have these two weaknesses

Read more

Summary

Introduction

Random fields (RFs) have been successfully applied, as a statistical model, to the analysis of biological sequences, text and image processing, as well as many areas of computer vision and artificial intelligence. Some RFs have been defined by multivariate Skew-Normal (SN) distribution (Azzalini, 1985) [5], multivariate Closed Skew-Normal (CSN) distribution (González-Farías et al, 2004) [10], multivariate Skew-t distribution and its general form, multivariate Extended Skew-t (EST) distribution (Arellano-Valle and Genton, 2010) [3]. We show that the RFs generated by some versions of multivariate SN, CSN and EST distributions have two weakness in application to the spatial interpolation.

Multivariate closed skew normal and extended skew-t distributions
Main results
54 Generalized Asymmetric Laplace random fields
The Generalized skew Laplace random fields
Application to spatial prediction
Discussion and Conclusion

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.