Open AccessModelling of GRDP the Construction Sector in Java Island Using Robust Geographically and Temporally Weighted Regression (RGTWR)
Author(s) -
Sugi Haryanto,
Muhammad Nur Aidi,
Anik Djuraidah
Publication year - 2019
Publication title -
international journal of scientific research in science, engineering and technology
Language(s) - English
Resource type - Journals
eISSN - 2395-1990
pISSN - 2394-4099
DOI - 10.32628/ijsrset196141
Subject(s) - outlier , estimator , statistics , regression analysis , java , linear regression , population , mathematics , econometrics , geography , computer science , demography , sociology , programming language
Infrastructure development is government focus in 2015-2019. One indicator used to measure economic activities construction sector in one area is GRDP of the construction sector. The Robust Geographically and Temporally Weighted Regression (RGTWR) model is the development of GTWR model to overcome the outliers. This study used GRDP the construction sector as a response variable with population, local revenue, area, and the number of construction establishments as explanatory variables. The RGTWR model is more effective in describing the value of data GRDP the construction sector of the regencies/municipalities in Java Island in 2010-2016. This is indicated by a decrease in the value of RMSE, MAD, and MAPE. The RGTWR model with M-estimator has been able to reduce the measure goodness of fit model, even though the decrease is not too large. Some residuals the RGTWR model still detected as outlier so it is recommended to use another estimator.