Spatially Varying Coefficient Regression Models : Diagnostic and Remedial Methods for Collinearity

ISBN-13: 9783639114379 / Angielski / Miękka / 2009 / 132 str.

David Wheeler

Opis

The realization in the statistical and geographicalsciences that relationships between explanatoryvariables and a response variable in a regressionmodel may vary across a study area has lead to thedevelopment of regression models with spatiallyvarying coefficients. Two such models aregeographically weighted regression and Bayesianregression models with spatially varyingcoefficients. In the application of these models,inference on the regression coefficient spatialprocesses is typically of primary interest. Thepresence of collinearity necessitates the use ofdiagnostic tools in local regression model buildingto highlight areas in which the results are notreliable for statistical inference, in addition topresenting an opportunity for remedial methods. Thisbook, therefore, provides diagnostic tools andremedial methods for spatially varying coefficientregression models and includes real-world andsimulated examples demonstrating the utility of thenew techniques. This book sheds light on the issuesof implementing and interpreting results from thesemodels and should prove especially useful to spatialdata analysts in Geography, Statistics, and PublicHealth.

O autorze

David C Wheeler, PhD, MAS, MA: Studied Statistics and Geography
at The Ohio State University. Pursuing Master s in Public Health
at Harvard University. Cancer Prevention Fellow at the National
Cancer Institute, Rockville, MD.