SYIFA FADIA SALSABILA, . (2022) PEMODELAN MORTALITAS PENDERITA DIFTERI DI INDONESIA MENGGUNAKAN REGRESI ZERO-INFLATED NEGATIVE BINOMIAL. Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.
![[img]](http://repository.unj.ac.id/style/images/fileicons/text.png) |
Text
COVER.pdf Download (1MB) |
|
Text
BAB 1.pdf Download (617kB) |
|
Text
BAB 2.pdf Restricted to Registered users only Download (697kB) |
|
Text
BAB 3.pdf Restricted to Registered users only Download (700kB) |
|
Text
BAB 4.pdf Restricted to Registered users only Download (718kB) |
|
Text
BAB 5.pdf Restricted to Registered users only Download (417kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (554kB) |
|
Text
LAMPIRAN.pdf Restricted to Registered users only Download (959kB) |
Abstract
Indonesia merupakan salah satu negara penyumbang kasus kejadian difteri tertinggi di dunia yang masih disertai dengan kematian. Penelitian ini bertujuan untuk memodelkan jumlah mortalitas difteri di Indonesia dan mengetahui faktor-faktor yang signifikan mempengaruhinya. Data yang dianalisis dalam penelitian ini mengindikasikan terjadinya overdispersi yang disebabkan oleh excess zeros karena pada variabel respon, yaitu data jumlah mortalitas difteri, terdapat 73% pengamatan bernilai nol dan persentase nilai nol pada data lebih besar dari persentase nilai nol berdasarkan perhitungan distribusi Poisson sehingga regresi Poisson menjadi kurang sesuai dalam memodelkan data tersebut. Model alternatif yang mampu mengatasi permasalahan overdispersi yang disebabkan oleh excess zeros adalah regresi Zero-Inflated Negative Binomial (ZINB). Berdasarkan kriteria uji Vuong dan nilai AIC, model regresi ZINB memiliki performa yang relatif lebih cocok digunakan daripada regresi Poisson. Berdasarkan pengujian hipotesis parameter model regresi ZINB, pada model count diperoleh kelima variabel penjelas yang digunakan dalam penelitian ini berpengaruh signifikan terhadap jumlah mortalitas difteri di Indonesia, sementara pada model logit tidak ada variabel penjelas yang signifikan. ******************************************* Indonesia is one of the countries that contribute to the highest incidence of diphtheria cases in the world which is still accompanied by death. This study aims to model the number of diphtheria mortality in Indonesia and to determine the factors that significantly influence it. The data analyzed in this study indicates the occurrence of overdispersion caused by excess zeros because the response variable, namely the data on the number of diphtheria mortality, there are 73% of observations with zero values and the percentage of zero values in the data is greater than the percentage of zero values based on the calculation of the Poisson distribution so that the Poisson regression is not suitable for modeling the data. An alternative model that is able to overcome the overdispersion caused by excess zero is the Zero-Inflated Negative Binomial (ZINB) regression. Based on the Vuong test and the AIC value criteria, the ZINB regression model has a relatively more suitable performance than the Poisson regression. Based on the hypothesis testing of the ZINB regression model parameters, the count model obtained five explanatory variables used in this study that had a significant effect on the number of diphtheria mortality in Indonesia, while in the logit model there is no significant explanatory variable
| Item Type: | Thesis (Sarjana) |
|---|---|
| Subjects: | Sains > Matematika |
| Divisions: | FMIPA > S1 Statistika |
| Depositing User: | PKL . |
| Date Deposited: | 19 Nov 2024 04:27 |
| Last Modified: | 19 Nov 2024 04:27 |
| URI: | http://repository.unj.ac.id/id/eprint/52151 |
Actions (login required)
 |
View Item |
