 |
 |
Dear All > Examining Spatial Heterogeneity through Geographically Weighted Regression > > <> > Prof . Stewart Fotheringham, NUI . > March 7th (Wednesday) ,15:00, Room A2002, Main Building > > > A common statistical modelling technique used in spatial analysis is > that of regression. In standard applications of regression a dependent > variable is linked to a set of independent variables with one of the > main outputs of the modelling procedure being the estimation of a > parameter that links each independent variable to the dependent > variable. A major problem with this technique when applied to spatial > data is that the processes being examined are assumed to be constant > over space - that is, one model fits all. Geographically Weighted > Regression (GWR) is a statistical technique that allows the modelling > of processes that vary over space. GWR results in a set of local > parameter estimates for each relationship which can be mapped to > produce a parameter surface across the study region. In this way, GWR > can provide information on the nature of the processes being > investigated and is increasingly preferred to traditional global types > of regression modelling. The seminar will provide a general overview > to GWR as well as introducing some of the statistical issues we are > currently grappling with. The use of user-friendly software for GWR > will also be described. > Tea is taken about 15 mins before the talk All Welcome ! Best Gilbert -- _____________________________ Prof. Gilbert MacKenzie Dept. of Mathematics & Statistics, University of Limerick, Limerick Ireland CoB ~ http://www.ul.ie/biostatistics ISA ~ http://www.istat.ie. Gilbert ~ http://www.staff.ul.ie/mackenzieg Email: [log in to unmask] Tel: +353 (0)61 213499 Fax: +353 (0)61 334927 _________________________