Pemodelan Jumlah Kematian Di Jawa Timur Ibu Dangan Pendekatan Generalized Poisson Regression (GPR) Dan Regresi Binomial Negatif

Pernama, Retdiansyah Risky Angga (2014) Pemodelan Jumlah Kematian Di Jawa Timur Ibu Dangan Pendekatan Generalized Poisson Regression (GPR) Dan Regresi Binomial Negatif. Other thesis, Insititut Teknologi Sepuluh Nopember.

Text 1312105009_Undergraduate_Thesis.pdf Download (1MB)

Abstract

Data jumlah kematian ibu dalam penelitian ini merupakan data diskrit (count) yang mempunyai varians lebih besar dibandingkan rata- ratanya (overdispersi). Untuk menangani masalah overdispersi, dapat dilakukan pemodelan dengan GPR dan Regresi Binomial Negatif. Model terbaik GPR menghasilkan 4 variabel prediktor yang signifikan mempengaruhi jumlah kasus kematian ibu tiap kabupaten/kota di Jawa Timur antara lain persentase ibu hamil yang melaksanakan program K1, persentase ibu hamil yang melaksanakan program K4, persentase ibu hamil yang mendapatkan Fe1 dan persentase persalinan ibu hamil yang mendapatkan Fe3. Sedangkan model terbaik menggunakan regresi binomial negatif menghasilkan 4 variabel prediktor yang signifikan yaitu terdiri atas persentase ibu hamil yang melaksanakan program K4, persentase ibu nifas yang mendapatkan vitamin A, persentase persalinan ditolong oleh tenaga kesehatan, dan persentase ibu hamil yang mendapatkan Fe3. Model GPR menghasilkan nilai AIC sebesar 293,2 dan model regresi binomial negatif menghasilkan nilai AIC sebesar 289,28. Maka model terbaik diperoleh dari model regresi binomial negatif karena menghasilkan nilai AIC terkecil.
=============================================================================================================================
Data the number of maternal mortality in this study is an example of discrete data (count) which have variance greater than the mean or called overdispersion. To deal with overdispersion, do modeling with Generalized Poisson Regression (GPR) and Negative Binomial Regression. The best model of GPR produces a significant predictor variables 4 affects the number of maternal mortality in in East Java such as, the percentage of pregnant women who carry out the program of K1, the percentage of pregnant women who carry out program K4, the percentage of pregnant women who get the Fe1 and percentage of pregnant women who get the Fe3. The best model of negative binomial regression produces a significant predictor variables 4 affects the number of maternal mortality in in East Java such as percentage of pregnant women who carry out program K4, the percentage of mothers who received vitamin A parturition, childbirth rescued by percentage of health care personnel, and the percentage of pregnant women who get the Fe3. Model GPR produces a value of AIC of binomial regression model and the negative binomial regression model produces AIC value of 293,28 . The best model is obtained from a negative binomial regression models because it produces the smallest of AIC value.

Item Type:	Thesis (Other)
Additional Information:	RSSt 519.536 Per p-1, 2014
Uncontrolled Keywords:	AIC, Generalized Poisson Regression, Kematian Ibu, Regresi Binomial Negatif, eneralized Poisson Regression, Maternal Mortality, Negative Binomial Regression
Subjects:	Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Mr. Marsudiyana -
Date Deposited:	08 Jan 2024 03:31
Last Modified:	08 Jan 2024 03:35
URI:	http://repository.its.ac.id/id/eprint/105393

Actions (login required)

View Item