The use of GIS and Cluster analysis to enhance property valuation modelling in rural Victoria which was published in the Journal of Spatial Science.
Rural property valuation is typically performed using manual techniques in Victoria, Australia. Regression analysis is a semi‐automated approach to statistically define numerical models for property valuation, however there can be some bias in regression models developed without the determination of sub‐markets (Watkins, 1998). Cluster analysis is a data driven technique to create sub‐market groups that may help overcome this bias.
This paper reports on research which seeks to ascertain if a higher level of price estimation can be obtained using property segmented using cluster analysis compared to sub‐markets created using geographical ‘a‐priori’ techniques; in this research the Local Government Area to which each property belongs. The research has shown that an increase has occurred for one of the sets of models which used cluster analysis, however it is envisioned that automated techniques to model rural property values would at this stage not be used as the sole valuation technique even if the accuracy of the models were to be improved dramatically.