Abstract

The use of ordinary linear regression model in spatial heterogeneity data often does not suitable within the data points, especially the relationship between response variable and explanatory variables. Therefore, the geographically weighted t regression (GWtR) is used to overcome spatial heterogeneity term. The model is an extension of geographically weighted regression (GWR) which the response variable follows multivariate t distribution. The aim of this study is to obtain the estimator of geographically weighted multivariate t regression (GWMtR) model with known degrees of freedom. The maximum likelihood estimation (MLE) method will be applied to maximize a weighted logarithm likelihood function. Based on the EM algorithm, the estimator of geographically weighted multivariate t regression model can be determined.

Original languageEnglish
Pages (from-to)45-51
Number of pages7
JournalJournal of Theoretical and Applied Information Technology
Volume92
Issue number1
Publication statusPublished - Oct 2016

Keywords

  • EM algorithm
  • Geographically weigted regression
  • Maximum likelihood estimation (MLE)
  • Multivariate t model

Fingerprint

Dive into the research topics of 'Parameter estimation of geographically weighted multivariate t regression model'. Together they form a unique fingerprint.

Cite this