PEMODELAN INTERPOLASI KRIGING PADA DISTRIBUSI KERNEL BISQUARE-GWR UNTUK NIR BANDUNG TIMUR

• Main Author: Amanda Widiareni Putri, Dea
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39874
• Institution: Institut Teknologi Bandung

Nilai Indikasi Rata-rata (NIR) which usually reffered as lang price can be estimated to assess the usefullness of land are in human life which leads to determine a state revenue the from tax sector in the form of Pajak Bumi dan Bangunan (PBB). In this final assignment research, the estimation land price in East Bandung Region is carried out by calculating the proximity of land polygon field to public facilities around it. Estimation are made of regression equation by assigning geographical weight values that take focus on a local area, that is Geographically Weighted Regression (Fotheringham, Brunsdon, & Charlton, 2002). The Geographically Weighted Regression (GWR) mehod used in this final project is adaptive spatial kernel function with bisquare weighting. The result are a number of object coverage limits which affected in each land price expressed in the form of integer, namely optimum bandwidth. The optimum bandwidth of all sample data is then interpolated to estimated bandwidth value of all polygon field in the entire study area using kriging method. The bandwidth model for all sample point has 33.011 rupiah/m2 for RMSE value, while the bandwidth model for all test point of kriging interpolation method is 114.600 rupiah/m2. It shows that kriging interpolation method decrase the accuracy of land price bandwidth model.