PENGEMBANGAN MODEL REGRESI TERBOBOTI SECARA GEOGRAFIS DAN TEMPORAL Studi Kasus: Angka Insiden Demam Berdarah Dengue di Surabaya

BAHARUDDIN, 091170203 (2015) PENGEMBANGAN MODEL REGRESI TERBOBOTI SECARA GEOGRAFIS DAN TEMPORAL Studi Kasus: Angka Insiden Demam Berdarah Dengue di Surabaya. Disertasi thesis, UNIVERSITAS AIRLANGGA.

[img]
Preview
Text (ABSTRAK)
gdlhub-gdl-s3-2015-baharuddin-35727-6.abstr-t.pdf

Download (194kB) | Preview
[img] Text (FULL TEXT)
gdlhub-gdl-s3-2015-baharuddin-35727-FULLt.pdf
Restricted to Registered users only

Download (1MB) | Request a copy
Official URL: http://lib.unair.ac.id

Abstract

One of the problems associated with regression modeling on the spatial and temporal data is that parameter of the model is not the same for each location and/or time. Another problem is that modeling on the time series data usually results temporally correlated errors. In this study, three regression models have extended to solve the problems, those are: geographically weighted regression (GWR) with autocorrelation, temporally weighted regression (TWR) with autocorrelation, and geographically and temporally weighted regression (GTWR). The purposes of this study are (1) to gain the parameter estimations of GWR, TWR, and GTWR models; (2) to get the statistic in testing the accuracy of the GWR,TWR, and GTWR models; (3) to analyze the spatial and/or temporal relationships between dengue fever incidence and climatic factors in Surabaya.The statistic for testing the goodness of fit of the GWR, TWR, and GTWR models is based on F-statistic by comparing its significance with global regression model. Meanwhile, the significance test for variation of each set of k  parameters is based on k F -statistic. Dengue fever incidence in Surabaya has a correlation with relative humidity and minimum atmospheric pressure. The relationship between dengue fever incidence and relative humidity is spatially different, while the effect of minimum atmospheric pressure is constant for all study regions. The relationships between dengue fever incidence and both climatic factors are temporally different. Spatially and temporally, the relationships between dengue fever incidence and both climatic variables are non-stationary for each district and month.

Item Type: Thesis (Disertasi)
Additional Information: KKC KK DIS M 04/15 Bah p
Uncontrolled Keywords: geographically and temporally weighted regression, dengue fever,Surabaya
Subjects: G Geography. Anthropology. Recreation > G Geography (General) > G1-922 Geography (General)
H Social Sciences > HB Economic Theory > HB1951 Population geography
R Medicine > RC Internal medicine > RC109-216 Infectious and parasitic diseases
Divisions: 09. Sekolah Pasca Sarjana > Ilmu Matematika & IPA
Creators:
CreatorsEmail
BAHARUDDIN, 091170203UNSPECIFIED
Contributors:
ContributionNameEmail
ContributorSuhariningsih, Prof.,Dr.,IrUNSPECIFIED
ContributorBrodjol Sutijo Suprih Ulama, Dr.,M.SiUNSPECIFIED
Depositing User: Nn Anisa Septiyo Ningtias
Last Modified: 01 Sep 2016 07:22
URI: http://repository.unair.ac.id/id/eprint/32784
Sosial Share:

Actions (login required)

View Item View Item