Open AccessOptimum scoring scheme to classify villages into urban-rural group
Author(s) -
R. P. Manik,
. Herlina,
Aji Hamim Wigena,
Soedarti Surbakti,
Bagus Sartono
Publication year - 2019
Publication title -
iop conference series. earth and environmental science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.179
H-Index - 26
eISSN - 1755-1307
pISSN - 1755-1315
DOI - 10.1088/1755-1315/299/1/012026
Subject(s) - constraint (computer aided design) , statistics , rural area , raw data , computer science , mathematics , medicine , geometry , pathology
The scoring method has been used by BPS-Statistics Indonesia since the 1980s to classify urban/rural areas. Currently, the 2010 scoring method uses total score 10 as a threshold to classify villages to urban-rural status. If the total score more than or equal 10, the villages are classified as urban, and rural otherwise. Applying the 2010 scoring method on raw data of Pendataan Potensi Desa (PODES) 2008 and 2014 shows 1266 villages change from urban to rural. Therefore, it is necessary to evaluate the determinant of predictors and the criteria of each predictor. The purpose of this research is to show the optimum scoring method from several optimizations that change the predictors and several optimizations that add new predictors. Exploratory Data Analysis (EDA) used to obtain the predictors and scores for each new criterion. In relation to this research problem, optimization is used to get the best results under given constraint. The constraint of the optimization carried out is the assumption that the changes in rural to urban status are increasing, and the changes in urban to rural are not existing. The optimum scoring method obtained from this study is the one excluding cinema (X8), changing the criteria of percentage of households with cable phone (X11) and percentage of households with electricity (X12), replacing predictor hotels (X10) into a starred hotel and adding minimarket as a new predictor. This optimization uses 12 variables with threshold 10. The implication of this study for future research is the use of more advanced statistical methods than EDA to determine the criteria of each predictor.