Geographically weighted regression with a non-Euclidean distance metric: A case study using hedonic house price data

Binbin Lu, Martin Charlton, Paul Harris, Stewart Fotheringham

Research output: Contribution to journalArticle

73 Citations (Scopus)

Abstract

Geographically weighted regression (GWR) is an important local technique for exploring spatial heterogeneity in data relationships. In fitting with Tobler's first law of geography, each local regression of GWR is estimated with data whose influence decays with distance, distances that are commonly defined as straight line or Euclidean. However, the complexity of our real world ensures that the scope of possible distance metrics is far larger than the traditional Euclidean choice. Thus in this article, the GWR model is investigated by applying it with alternative, non-Euclidean distance (non-ED) metrics. Here we use as a case study, a London house price data set coupled with hedonic independent variables, where GWR models are calibrated with Euclidean distance (ED), road network distance and travel time metrics. The results indicate that GWR calibrated with a non-Euclidean metric can not only improve model fit, but also provide additional and useful insights into the nature of varying relationships within the house price data set.

Original languageEnglish (US)
Pages (from-to)660-681
Number of pages22
JournalInternational Journal of Geographical Information Science
Volume28
Issue number4
DOIs
StatePublished - Apr 2014
Externally publishedYes

Fingerprint

regression
Travel time
travel time
road network
price
travel
geography
Law

Keywords

  • local regression
  • non-stationarity
  • real estate
  • road network distance
  • travel time

ASJC Scopus subject areas

  • Information Systems
  • Geography, Planning and Development
  • Library and Information Sciences

Cite this

Geographically weighted regression with a non-Euclidean distance metric : A case study using hedonic house price data. / Lu, Binbin; Charlton, Martin; Harris, Paul; Fotheringham, Stewart.

In: International Journal of Geographical Information Science, Vol. 28, No. 4, 04.2014, p. 660-681.

Research output: Contribution to journalArticle

@article{fbdc4e53de4440ef8ad53e82eb0377e6,
title = "Geographically weighted regression with a non-Euclidean distance metric: A case study using hedonic house price data",
abstract = "Geographically weighted regression (GWR) is an important local technique for exploring spatial heterogeneity in data relationships. In fitting with Tobler's first law of geography, each local regression of GWR is estimated with data whose influence decays with distance, distances that are commonly defined as straight line or Euclidean. However, the complexity of our real world ensures that the scope of possible distance metrics is far larger than the traditional Euclidean choice. Thus in this article, the GWR model is investigated by applying it with alternative, non-Euclidean distance (non-ED) metrics. Here we use as a case study, a London house price data set coupled with hedonic independent variables, where GWR models are calibrated with Euclidean distance (ED), road network distance and travel time metrics. The results indicate that GWR calibrated with a non-Euclidean metric can not only improve model fit, but also provide additional and useful insights into the nature of varying relationships within the house price data set.",
keywords = "local regression, non-stationarity, real estate, road network distance, travel time",
author = "Binbin Lu and Martin Charlton and Paul Harris and Stewart Fotheringham",
year = "2014",
month = "4",
doi = "10.1080/13658816.2013.865739",
language = "English (US)",
volume = "28",
pages = "660--681",
journal = "International Journal of Geographical Information Science",
issn = "1365-8816",
publisher = "Taylor and Francis Ltd.",
number = "4",

}

TY - JOUR

T1 - Geographically weighted regression with a non-Euclidean distance metric

T2 - A case study using hedonic house price data

AU - Lu, Binbin

AU - Charlton, Martin

AU - Harris, Paul

AU - Fotheringham, Stewart

PY - 2014/4

Y1 - 2014/4

N2 - Geographically weighted regression (GWR) is an important local technique for exploring spatial heterogeneity in data relationships. In fitting with Tobler's first law of geography, each local regression of GWR is estimated with data whose influence decays with distance, distances that are commonly defined as straight line or Euclidean. However, the complexity of our real world ensures that the scope of possible distance metrics is far larger than the traditional Euclidean choice. Thus in this article, the GWR model is investigated by applying it with alternative, non-Euclidean distance (non-ED) metrics. Here we use as a case study, a London house price data set coupled with hedonic independent variables, where GWR models are calibrated with Euclidean distance (ED), road network distance and travel time metrics. The results indicate that GWR calibrated with a non-Euclidean metric can not only improve model fit, but also provide additional and useful insights into the nature of varying relationships within the house price data set.

AB - Geographically weighted regression (GWR) is an important local technique for exploring spatial heterogeneity in data relationships. In fitting with Tobler's first law of geography, each local regression of GWR is estimated with data whose influence decays with distance, distances that are commonly defined as straight line or Euclidean. However, the complexity of our real world ensures that the scope of possible distance metrics is far larger than the traditional Euclidean choice. Thus in this article, the GWR model is investigated by applying it with alternative, non-Euclidean distance (non-ED) metrics. Here we use as a case study, a London house price data set coupled with hedonic independent variables, where GWR models are calibrated with Euclidean distance (ED), road network distance and travel time metrics. The results indicate that GWR calibrated with a non-Euclidean metric can not only improve model fit, but also provide additional and useful insights into the nature of varying relationships within the house price data set.

KW - local regression

KW - non-stationarity

KW - real estate

KW - road network distance

KW - travel time

UR - http://www.scopus.com/inward/record.url?scp=84894273575&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84894273575&partnerID=8YFLogxK

U2 - 10.1080/13658816.2013.865739

DO - 10.1080/13658816.2013.865739

M3 - Article

AN - SCOPUS:84894273575

VL - 28

SP - 660

EP - 681

JO - International Journal of Geographical Information Science

JF - International Journal of Geographical Information Science

SN - 1365-8816

IS - 4

ER -