Further Evaluating the Impact of Kernel and Bandwidth Specifications of Geographically Weighted Regression on the Equity and Uniformity of Mass Appraisal Models

Research has consistently demonstrated that geographically weighted regression (GWR) models significantly improve upon accuracy of ordinary least squares (OLS)-based computer-assisted mass appraisal (CAMA) models by more accurately accounting for the effects of location (Fotheringham et al. 2002; LeSage 2004; Huang et al. 2010). Bidanset and Lombard (2014a, 2017) previously studied the impacts of various kernel and bandwidth combinations employed in building residual (i.e. sale price less land value) GWR CAMA models and found that the specification of each does bear significant effect on valuation equity attainment. This paper builds upon the previous research by comparing performance of weighting specifications of non-building residual (i.e. full sale price) GWR CAMA models using new data of a different geographic real estate market. We find that the exponential kernel and fixed bandwidth together achieve a superior COD for our data, and that COD does fluctuate depending on the GWR weighting specification.

[1]  Michael Ball,et al.  Recent Empirical Work on the Determinants of Relative House Prices , 1973 .

[2]  Robert S. Bednarz,et al.  A Hedonic Model of Prices and Assessments for Single-Family Homes: Does the Assessor Follow the Market or the Market Follow the Assessor? , 1975 .

[3]  N. Sugiura.  Further analysts of the data by akaike' s information criterion and the finite corrections , 1978 .

[4]  Luc Anselin,et al.  Do spatial effects really matter in regression analysis , 2005 .

[5]  W. Cleveland,et al.  Locally Weighted Regression: An Approach to Regression Analysis by Local Fitting , 1988 .

[6]  Daniel P. McMillen,et al.  One Hundred Fifty Years of Land Values in Chicago: A Nonparametric Approach , 1996 .

[7]  M. Charlton,et al.  Geographically Weighted Regression: A Natural Evolution of the Expansion Method for Spatial Data Analysis , 1998 .

[8]  Martin Charlton,et al.  Spatial Nonstationarity and Autoregressive Models , 1998 .

[9]  J. LeSage.  A Family of Geographically Weighted Regression Models , 2004 .

[10]  L. Zhang,et al.  Comparison of bandwidth selection in application of geographically weighted regression : a case study , 2008 .

[11]  William J. McCluskey,et al.  Using Geographically Weighted Regression to Detect Housing Submarkets: Modeling Large-Scale Spatial Variations in Value , 2008 .

[12]  D. Lambert,et al.  Geographically weighted regression bandwidth selection and spatial autocorrelation: an empirical example using Chinese agriculture data , 2010 .

[13]  Bo Wu,et al.  Geographically and temporally weighted regression for modeling spatio-temporal variation in house prices , 2010, Int. J. Geogr. Inf. Sci..

[14]  A. Stewart Fotheringham,et al.  Geographically Weighted Regression: A Method for Exploring Spatial Nonstationarity , 2010 .

[15]  D. McMillen,et al.  Estimation and Hypothesis Testing for Nonparametric Hedonic House Price Functions , 2010 .

[16]  J. Wayne Moore,et al.  Using geographic-attribute weighted regression for CAMA modeling , 2010 .

[17]  Tony Lockwood,et al.  Efficacy in Modelling Location Within the Mass Appraisal Process , 2011 .

[18]  David McIlhatton,et al.  Prediction accuracy in mass appraisal: a comparison of modern approaches , 2013 .

[19]  Martin Charlton,et al.  The GWmodel R package: further topics for exploring spatial heterogeneity using geographically weighted models , 2013, Geo spatial Inf. Sci..

[20]  Paul E. Bidanset,et al.  The effect of kernel and bandwidth specification in geographically weighted regression models on the accuracy and uniformity of mass real estate appraisal , 2014 .

[21]  Paul E. Bidanset,et al.  Evaluating Spatial Model Accuracy in Mass Real Estate Appraisal: A Comparison of Geographically Weighted Regression and the Spatial Lag Model , 2014 .