کد مقاله | کد نشریه | سال انتشار | مقاله انگلیسی | نسخه تمام متن |
---|---|---|---|---|
1064582 | 1485798 | 2012 | 13 صفحه PDF | دانلود رایگان |
Historically, major contributions to popularizing spatial statistics derived from the pioneering work of Cliff and Ord. One outcome was the development of spatial econometrics. With the passing of time, this body of work merged with geostatistics to form the present day discipline of spatial statistics. The families of auto- and semivariogram models constitute a prominent component of the subject matter of contemporary spatial statistics. Its expansion from linear to generalized linear statistical models involves new methodologies, one of which is eigenvector spatial filtering. This paper presents evidence that this particular new methodology furnishes an effective dimension reduction substitution for the spatial lag matrix appearing in spatial auto-models. It also summarizes ongoing extensions of this methodology to space-time and spatial interaction data. Eigenvector spatial filtering methodology presents a new frontier for spatial statistical research.
Journal: Spatial Statistics - Volume 1, May 2012, Pages 3–15