Pemodelan Kasus Pneumonia pada Balita di Jawa Tengah Menggunakan Multivariate Adaptive Generalized Poisson Regression Spline (MAGPRS)

Nisa, Dinda Khairun (2023) Pemodelan Kasus Pneumonia pada Balita di Jawa Tengah Menggunakan Multivariate Adaptive Generalized Poisson Regression Spline (MAGPRS). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211940000022-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (6MB) | Request a copy

Abstract

Pneumonia masih menjadi penyebab utama kematian terbanyak pada balita di Indonesia setelah diare dengan persentase 9,4%. Provinsi yang memiliki jumlah kematian balita tertinggi akibat pneumonia di Indonesia tahun 2021 adalah Jawa Tengah dengan angka prevalensi mencapai 42,53%. Melihat tingginya jumlah kematian balita akibat pneumonia di Provinsi Jawa Tengah, maka diperlukan penelitian untuk menekan kasus pneumonia balita dengan pemodelan jumlah kasus pneumonia pada balita di Jawa Tengah. Tujuan penelitian ini adalah untuk mendeskripsikan kasus pneumonia balita di Jawa Tengah dan faktor-faktor yang diduga memengaruhinya dan menentukan model terbaik pada jumlah kasus pneumonia balita di Provinsi Jawa Tengah dengan metode Multivariate Adaptive Generalized Poisson Regression Spline (MAGPRS). Adapun dasar teori yang digunakan yaitu Regresi Poisson, Generalized Poisson Regression, Multivariate Adaptive Regression Spline (MARS), MAGPRS, dan penelitian terdahulu. Data yang digunakan dalam penelitian ini merupakan data sekunder dari Badan Pusat Statistik Provinsi Jawa Tengah dan Profil Kesehatan Provinsi Jawa Tengah 2021. Penelitian ini menggunakan metode MARS dengan estimator Generalized Poisson, sehingga menjadi metode MAGPRS karena data penelitian tidak menunjukkan pola tertentu dan data mengalami kasus overdispersi. Model MAGPRS terbaik ditentukan berdasarkan kriteria Generalized Cross Validation (GCV) minimum dan diperoleh model terbaik dari hasil analisis adalah model dengan Basis Function (BF), Maximum Interaction (MI), dan Minimum Observation (MO) berturut-turut sebesar 32, 3, dan 1 dengan nilai GCV sebesar 788,183. Model tersebut mampu menjelaskan 99,927% keragaman data yang ada. Variabel prediktor yang paling berpengaruh terhadap model adalah variabel persentase penduduk usia 35-44 tahun yang merokok, kepadatan penduduk per km2, persentase balita yang mendapat vitamin A, dan persentase Berat Badan Lahir Rendah (BBLR). Sementara itu, persentase balita gizi kurang tidak memiliki tingkat kepentingan variabel relatif pada model MAGPRS terpilih karena tidak masuk ke dalam model dan sudah terwakili oleh variabel lainnya.
=================================================================================================================================
Pneumonia remains the leading cause of death among toddlers in Indonesia after diarrhea, with a prevalence rate of 9.4%. The province with the highest number of toddler deaths due to pneumonia in Indonesia in 2021 was Central Java, with a prevalence rate of 42.53%. Given the high number of toddler deaths due to pneumonia in Central Java Province, research is needed to reduce pneumonia cases among toddlers by modeling the number of pneumonia cases in toddlers in Central Java. The objective of this research is to describe pneumonia cases among toddlers in Central Java and the suspected factors influencing them, as well as determining the best model for the number of pneumonia cases in toddlers in Central Java Province using the Multivariate Adaptive Generalized Poisson Regression Spline (MAGPRS) method. The theoretical basis used includes Poisson Regression, Generalized Poisson Regression, Multivariate Adaptive Regression Spline (MARS), MAGPRS, and previous studies. The data used in this research are secondary data from the Central Java Provincial Central Bureau of Statistics and the Health Profile of Central Java Province in 2021. This research utilizes the MARS method with the Generalized Poisson estimator, resulting in the MAGPRS method, as the research data does not exhibit a specific pattern and the data experiences overdispersion. The best MAGPRS model is determined based on the minimum Generalized Cross Validation (GCV) criteria. The best model obtained from the analysis consists of Basis Function (BF), Maximum Interaction (MI), and Minimum Observation (MO) with values of 32, 3, and 1, respectively, and a GCV value of 788.183. The model is able to explain 99.927% of the existing data variation. The most influential predictor variables in the model are the percentage of the population aged 35-44 years who smoke, population density per km2, the percentage of toddlers receiving vitamin A, and the percentage of Low Birth Weight (LBW) babies. Meanwhile, the percentage of malnourished
toddlers does not have a significant relative importance level in the selected MAGPRS model as it is not included in the model and is already represented by other variables.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Central Java, Generalized Poisson, MAGPRS, Toddler, Pneumonia, Balita, Jawa Tengah
Subjects:	R Medicine > RC Internal medicine > RC771 Pneumonia.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Dinda Khairun Nisa
Date Deposited:	27 Sep 2023 02:00
Last Modified:	27 Sep 2023 02:00
URI:	http://repository.its.ac.id/id/eprint/104459

Actions (login required)

View Item