Pemodelan Multivariate Adaptive Generalized Poisson Regression Spline Pada Kasus Jumlah Kematian Ibu Di Provinsi Jawa Timur

Prastika, Euodia Putri (2020) Pemodelan Multivariate Adaptive Generalized Poisson Regression Spline Pada Kasus Jumlah Kematian Ibu Di Provinsi Jawa Timur. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 06211640000026-Undergraduate_Thesis.pdf]
Preview
Text
06211640000026-Undergraduate_Thesis.pdf

Download (1MB) | Preview

Abstract

Ibu berperan penting dalam regenerasi manusia, sehingga keberadaan ibu menjadi perhatian pemerintah. Salah satu masal ah besar yang dialami oleh negara-negara, termasuk Indonesia adalah kematian ibu. Indonesia memiliki angka kematian ibu (AKI) tertinggi kedua di ASEAN, dimana tingginya AKI Indonesia merupakan sumbangsih dari AKI provinsi-provinsi di Indonesia, termasuk Provinsi Jawa Timur. Provinsi Jawa Timur masuk dalam tiga besar AKI tertinggi di Indonesia dengan nilai 91,45 per 100.000 kelahiran hidup pada tahun 2018. Tingginya AKI berbanding positif dengan jumlah kematian ibu, artinya semakin tinggi AKI, maka semakin banyak jumlah kematian ibu. Salah satu upaya yang dapat dilakukan untuk mengurangi AKI adalah pemodelan faktor-faktor yang mempengaruhi jumlah kematian ibu dengan metode regresi. Penelitian ini menggunakan metode Multivariate Adaptive Regression Splines (MARS) dengan estimator generalized Poisson, sehingga menjadi metode Multivariate Adaptive Generalized Poisson Regression Splines (MAGPRS). Model terbaik dari hasil analisis MAGPRS adalah model dengan BF=28, MI=2, dan MO=2. Setelah dilakukan backward stepwise, fungsi basis dari model tersebut menjadi 24, dimana akan menyusun persamaan MAGPRS terbaik. Variabel prediktor yang paling berpengaruh terhadap model secara berurutan adalah variabel persentase ibu nifas mendapat vitamin A, persentase peserta aktif KB, dan persentase kunjungan ibu hamil K4.
================================================================================================================
Mothers play an important role in producing offspring, so the existence of the mother is the concern of the government. One big problem experienced by many countries, including Indonesia, is maternal mortality. Among ASEAN Countries, Indonesia has the second highest maternal mortality ratio (MMR), where the high MMR in Indonesia is contributed by the MMR of Indonesia’s provinces, including East Java Province. East Java Province was included in top three highest MMR in Indonesia with 91.45 maternal mortalities per 100,000 live births in 2018. High MMR is proportional to the number of maternal mortalities, signifying that the higher the MMR then the number of maternal mortalities is increasing as well. To reduce the MMR, an effort that can be done is modelling the factors that influence the number of maternal mortalities using the regression method. This research used the Multivariate Adaptive Regression Splines (MARS) method with generalized Poisson estimator, thus became the Multivariate Adaptive Generalized Poisson Regression Splines (MAGPRS) method. The best model of MAGPRS analysis is a model with BF = 28, MI = 2, and MO = 2. After backward stepwise, the basis function of the model changed to 24, which formed the best MAGPRS equation. The most influential predictor variables toward the model sequentially were the percentage of postpartum mothers got vitamin A, the percentage of active birth control participants, and the percentage of visitation to pregnant women K4.

Item Type: Thesis (Other)
Additional Information: RSSt 519.536 Pra p-1 • Prastika, Euodia Putri
Uncontrolled Keywords: BF, Jawa Timur, Jumlah Kematian Ibu, MAGPRS, MI, MO, Number of Maternal Mortality
Subjects: Q Science > QA Mathematics > QA278 Cluster Analysis. Multivariate analysis. Correspondence analysis (Statistics)
Divisions: Faculty of Mathematics, Computation, and Data Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Euodia Putri Prastika
Date Deposited: 26 Aug 2020 02:29
Last Modified: 21 Dec 2023 01:41
URI: http://repository.its.ac.id/id/eprint/81120

Actions (login required)

View Item View Item