Pemodelan Demam Berdarah Dengue di Surabaya dengan Pendekatan Mixed Geographically Weighted Poisson Regression
Abstract: Mixed Geographically
Weighted Poisson Regression (Mixed GWPR) model
is an expansion of GWPR. Mixed GWPR
model has locally influential predictor variable and
globally predictor variable
influential. Generally, Mixed
GWPR model is: ; i
= 1, 2,
… , n. The
local parameter estimation of Mixed GWPR model using estimation method
of GWPR model : and using estimation of global poisson regression to estimate global parameter
: . The data used
in the application of the model
Mixed GWPR is data on the number of dengue hemorrhagic fever patients in
23 sub-district in
Surabaya on 2011
as a response
variable. While predictor variable are
population density, the
number of population aged less
than 16 years,
the population of at least high school education, the percentage of
households living clean and healthy behavior, wiggler free numbers , the number
of health facilities and the percentage of
healthy. Based on
data analysis, six
locally influential predictor
variables and one globally influential predictor variables
are obtained. From that matter, it can be concluded that the number of health
facilities not-significantly influenced. This might occur due to the high
number of health facilities but not matched by the number of adequate health
facilities. While the percentage of households living clean and healthy
behavior non-significantly in one sub-district in Surabaya, population density,
the number of population aged less than 16
years, the population
of at least
high school education, wiggler free
numbers and the percentage of healthy significantly in 23 sub-district
in Surabaya.
Penulis: Anisa Muthiatul
Husnah, Suliyanto & Toha Saifudin
Kode Jurnal: jpmatematikadd130072
