The Modelling of Residential Sales Prices with Kriging Using Different Distance Metrics in Different Correlation Functions

Authors : Olgun AYDIN, Semra ERPOLAT TAŞABAT

Pages : 627-633

View : 14 | Download : 5

Publication Date : 2016-09-30

Article Type : Research Paper

Abstract :Modelling and estimating sales prices based on economical conditions is so important for housing sector. In this study, firstly; investigated economic variables effected mostly to housing sales prices and created kriging model for housing sales prices in Dubai. For effective kriging model, researchers need to select most effective correlation functions and optimum correlation parameters. To determine better correlation function structure, used Euclidian and Canberra distances for both Exponential and Gaussian correlation functions. Simulation studies were applied to obtain optimum correlation function parameters based on Maximum Likelihood Estimationinsert ignore into journalissuearticles values(MLE);  procedure. For detecting normality of reponse values, Focused Information Criteriainsert ignore into journalissuearticles values(FIC); was used. Based on Cross Validation criteria, selected best correlation function with best distance metrics.
Keywords : Kriging, Euclidian Distance, Canberra Distance, Housing Price, Economical Condition