Some Fast Methods for Fitting Some One-parameter Spatial Models | Science Publications

Journal of Mathematics and Statistics

R. J. Martin¹ 

关键词: Conditional autoregression;    geographical spatial models;    simultaneous autoregression;    Biological community;    abundance vector;    diversity profiles;    VAR models;   

DOI  :  10.3844/jmssp.2005.326.336

学科分类：社会科学、人文和艺术（综合）

来源: Science Publications

【 摘 要 】

It is common in geographic modelling to use a one-parameter spatial model to specify theinverse covariance matrix in terms of I-bW, for some known matrix W. Exact Gaussian maximumlikelihood estimation of b requires evaluation of the determinant of the covariance matrix. For largedata sets, this evaluation of the determinant can be slow and good approximations can be useful.Seventy regional configurations are used to consider some approximations to the determinant of I-bWthat are fast to evaluate, and their usefulness is compared.

【 授权许可】

Unknown