Pemodelan Kasus Pneumonia pada Balita di RSUD Haji Provinsi Jawa Timur Menggunakan Regresi Logistik Ordinal SMOTE-N

Putri, Alicia Kartika (2026) Pemodelan Kasus Pneumonia pada Balita di RSUD Haji Provinsi Jawa Timur Menggunakan Regresi Logistik Ordinal SMOTE-N. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003221021-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Pneumonia merupakan salah satu penyebab utama kematian balita di dunia, termasuk di Indonesia. Pada tahun 2024, Jawa Timur tercatat sebagai provinsi dengan kasus tertinggi kedua. Di RSUD Haji Provinsi Jawa Timur, terdapat 203 kasus pneumonia balita pada periode Januari–Juni 2025. Penyakit ini diklasifikasikan menjadi ringan, sedang, dan berat, yang masing-masing memerlukan penanganan berbeda. Penelitian ini bertujuan untuk mendeskripsikan karakteristik pasien, mengatasi ketidakseimbangan data menggunakan metode Synthetic Minority Over-sampling Technique-Nominal (SMOTE-N), serta mendapatkan model terbaik dan faktor-faktor yang memengaruhi tingkat keparahan pneumonia pada balita. Metode yang digunakan adalah regresi logistik ordinal dengan pendekatan SMOTE-N. Hasil analisis deskriptif menunjukkan sebagian besar balita termasuk ke dalam kategori pneumonia ringan, yaitu sebanyak 97 balita. Data awal menunjukkan ketidakseimbangan data dengan proporsi ringan 53,9%, sedang 33,9%, dan berat 12,2%. Untuk mengatasi hal tersebut, diterapkan metode SMOTE-N. Hasil analisis SMOTE-N diperoleh 20 variasi kombinasi distribusi dan nilai k-nearest neighbor. Hasil penelitian menunjukkan bahwa model 1 dengan distribusi data 97:61:50 dan nilai r = 7 (model A) dan r = 9 (model B) merupakan model terbaik dengan nilai AICc terendah, yaitu masing-masing sebesar 432,114 dan 431,162. Dari kedua model tersebut, faktor yang berpengaruh signifikan terhadap tingkat keparahan pneumonia pada balita adalah usia, jenis kelamin, leukosit, dan status imunisasi untuk model A, serta usia, jenis kelamin, dan status imunisasi untuk model B.
=======================================================================================================================================
Pneumonia is one of the leading causes of death among young children worldwide, including in Indonesia. In 2024, East Java was recorded as the province with the second highest number of cases. At RSUD Haji Provinsi Jawa Timur, there were 203 pediatric pneumonia cases during the period of January–June 2025. This disease is classified into mild, moderate, and severe categories, each requiring different treatment. This study aims to describe patient characteristics, address data imbalance using the Synthetic Minority Over-sampling Technique-Nominal (SMOTE-N) method, and obtain the best model and factors affecting the severity of pneumonia in children. This study used ordinal logistic regression combined with SMOTE-N approach. Descriptive analysis results showed that most children fell into the mild pneumonia category, namely 97 children. The initial data showed data imbalance with proportions of mild 53.9%, moderate 33.9%, and severe 12.2%. To address this, the SMOTE-N method was applied. The SMOTE-N analysis yielded 20 variation combinations of proportion and k-nearest neighbor values. The results showed that model 1 with data distribution 97:61:50 and r = 7 (model A) and r = 9 (model B) were the best model with the lowest AICc values, namely 432.114 and 431.162, respectively. From both models, the factors that significantly affected the severity of pneumonia in children were age, gender, leukocyte count, and immunization status for model A, and age, gender, and immunization status for model B.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Ketidakseimbangan Data, Pneumonia Balita, Regresi Logistik Ordinal,SMOTE-N, Tingkat Keparahan, Imbalanced Data, Ordinal Logistic Regression, Pediatric Pneumonia, SeverityLevel, SMOTE-N
Subjects:	H Social Sciences > HA Statistics > HA31.3 Regression. Correlation. Logistic regression analysis. Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression R Medicine > RC Internal medicine > RC771 Pneumonia.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Alicia Kartika Putri
Date Deposited:	31 Jul 2026 04:25
Last Modified:	31 Jul 2026 04:25
URI:	http://repository.its.ac.id/id/eprint/140419

Actions (login required)

View Item