Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Media Statistika

IMPLEMENTATION OF LOCALLY COMPENSATED RIDGE-GEOGRAPHICALLY WEIGHTED REGRESSION MODEL IN SPATIAL DATA WITH MULTICOLLINEARITY PROBLEMS (Case Study: Stunting among Children Aged under Five Years in East Nusa Tenggara Province) Fadliana, Alfi; Pramoedyo, Henny; Fitriani, Rahma
MEDIA STATISTIKA Vol 13, No 2 (2020): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.14710/medstat.13.2.125-135

Abstract

East Nusa Tenggara Province, according to the findings of 2013 Baseline Health Research and 2016 and 2017 Nutritional Status Surveys, was recorded as the province with the highest prevalence of stunting in Indonesia. Efforts should be made to formulate policies that are integrated with spatial aspects in order to reduce the prevalence of stunting. The LCR-GWR model approach is used by using locally compensated ridge, which were meant to adjusts to the effect of collinearity between predictor variables (i.e., the factors affecting the prevalence of stunting) in each area. Results of the analysis showed that factors affecting the prevalence of stunting in all districts/cities in East Nusa Tenggara Province are the percentage of children aged under five who were weighed ≥ 4 times, the percentage of children aged under five who receive complete basic immunization, the percentage of households consuming iodized salt, the percentage of households with decent source of drinking water and the real per capita expenditure. The analysis showed that LCR-GWR is able to produce a better model than the GWR model in overcoming local multicollinearity problems in stunting in East Nusa Tenggara Province, with lower RMSE value (0.0344) than the GWR RMSE model (3.8899).
INTERPOLASI KRIGING DALAM PEMODELAN GSTAR-SUR DAN GSTARX-SUR PADA SERANGAN HAMA PENGGEREK BUAH KOPI Henny Pramoedyo; Arif Ashari; Alfi Fadliana
MEDIA STATISTIKA Vol 13, No 1 (2020): Media Statistika
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (94.664 KB) | DOI: 10.14710/medstat.13.1.25-35

Abstract

The GSTAR and GSTARX models with the SUR approach normally can only be used in forecasting an event in the future in locations where the data is indeed used in forming the model. The problem that sometimes occurs in some cases is that not all locations that want to be modeled do not have data, or if there is data, the data is not as complete as other locations. This study uses GSTAR and GSTARX modeling with the SUR approach and combines them with the kriging interpolation technique in forecasting in an unobserved location. The case study used in this research is PBKo attack forecasting in Probolinggo Regency, where it is simulated that Watupanjang Village is an unobserved location because the location of coffee plantations in the area is difficult to reach due to difficult terrain / access roads. The results showed that PBKo pest attacks in the Probolinggo Regency could be predicted using the GSTAR model (1, [1,12]) and the GSTARX model (1, [1.12]) (10,0,0). Both models, both GSTAR Kriging and GSTARX Kriging, can be relied upon as an alternative to predicting PBKo pests in unobserved locations or where insufficient data are available.