An Introduction to Geographically Weighted Regression
The data we measure on our environment represent the outcomes of unknown spatial processes. We typically use spatial associations between observations on different variables to infer something about these processes. One of the most common ways to do this is to formulate a linear or non-linear model linking a dependent variable, y, to set of covariates, x1, x2...xn, and to estimate the parameters of this model by regression. A potential problem with this framework is that it assumes that the parameter estimates in the model (and hence the processes these parameters represent) are constant over space whereas in reality they might vary. Geographical Weighted Regression (GWR) is a statistical technique based to uncover potential spatial variations in the processes that produce the data we observe about the real world. This workshop will introduce GWR, both conceptually and statistically, and will demonstrate its use in both simulated and real datasets. Several features of GWR will also be explored and software for calibrating models by GWR will be introduced.
It will be assumed that participants will be familiar with regression but no knowledge of GWR will be assumed.
Presenter: Dr. Stewart Fotheringham, Professor, School of Geographical Sciences and Urban Planning, ASU
Date: February 17, 2017
Time: 9:00am - 4:00pm
Location: Cowden Bldg Room 124, Elinor Ostrom Lab
Fee: 0
Lunch will be included with this workshop
Register HERE