Kriging In Mass Appraisal For Rating

Abstract

The application of geostatistical method, kriging, to predict the unknown property values from observed data of known property values in the context of mass appraisal is described in this paper. This method uses semivariogram to define the weight of neighboring observations in the prediction procedure. Kriging has been used successfully in predicting underground surfaces such as ore bodies and soil nutrients but receive less attention from valuation and econometrics experts. Several variations of kriging are available but the focus of the paper is on ordinary kriging. The theory is described and computational method applicable to hedonic price equation is shown. Transaction data within Majlis Perbandaran Kulai from year 2004 to 2006 are used as a case study. The results are compared with the traditional method of Multiple Regression Analysis (MRA) in term of its prediction performance.