Invited Speakers

Professor A. Stewart Fotheringham, National University (Maynooth), Ireland

Geographically Weighted Regression and Geocomputation: An Overview of Recent Developments

Abstract: Geographically Weighted Regression (GWR) is a well-known technique to investigate spatial nonstationarity in relationships. It is part of a growing trend away from global ‘one-size-fits-all’ models and towards local models where relationships may vary over space. It is a quintessential geographical model because it uses locational information and produces geocoded outputs than can be mapped. The concept of GWR will be explained and some examples given of how it can be used to identify spatial variations in relationships. The talk will then describe some recent advances in GWR research including:

• The development of semi-parametric GWR models
• The development of new software for GWR, GWR 4.0
• Inference in GWR
• The transference of the concept of geographically weighting to other types of geographical models such as spatial interaction models

Biography: Professor Fotheringham is Director of the National Centre for Geocomputation at the National University of Ireland, Maynooth. He holds PhD and MA degrees from McMaster University in Canada and a BSc from Aberdeen University in Scotland. His research interests include: quantitative spatial analysis; spatial statistics; Geographic Information Science; spatial interaction modelling; spatial choice modelling; and spatial cognition. He has published 9 books; 28 chapters in books and over 100 refereed journal articles. Professor Fotheringham has been involved in large geographic Information initiatives in the USA, Canada, the UK and Ireland. He has received over 20 million euros in grants and is currently Director of a Staretgic Research Cluster in Advanced Geotechnologies. He has delivered over 100 presentations at academic conferences (many of them invited keynote addresses) and has given 60 invited seminars at universities around the world. He is one of the developers of Geographically Weighted Regression and has conducted 17 invited workshops on GWR.