Yustian, Ghevira Sandra Eka (2024) Penerapan Syntetic Minority Oversampling Technique (SMOTE) Pada Regresi Logistik Biner dalam Analisis Faktor-Faktor Kejadian Berat Badan Lahir Rendah (BBLR) di Kalimantan Selatan. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5003201012-Undergraduate-Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5003201012-Undergraduate-Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (3MB) | Request a copy |
Abstract
Angka Kematian Bayi (AKB) adalah jumlah kematian anak usia kurang dari 12 bulan per 1000 kelahiran hidup. AKB menjadi salah satu indikator kesehatan masyarakat yang mencerminkan kondisi hidup suatu negara. Dalam konteks AKB, dikenal istilah aturan 2/3 yang artinya 2/3 kematian bayi terjadi pada masa neonatal (usia 0-28 hari). Kematian neonatal ini mayoritas disebabkan karena kondisi Berat Badan Lahir Rendah (BBLR). Provinsi Kalimantan Selatan merupakan provinsi dengan AKB yang masih sangat tinggi yaitu 34 per 1000 kelahiran hidup di tahun 2022. Selain AKB yang masih tinggi, Provinsi Kalimantan Selatan juga memiliki persentase kejadian BBLR tertinggi di Indonesia yaitu 6,8% dan mengakibatkan 180 kematian neonatal. Penelitian ini bertujuan untuk menganalisis faktor risiko BBLR di Kalimantan Selatan dengan menggunakan model regresi logistik biner dan pendekatan Synthetic Minority Oversampling Technique (SMOTE) untuk mengatasi proporsi data yang tidak seimbang. Data yang digunakan berasal dari Survey Sosial Ekonomi Nasional (SUSENAS) 2022. Hasil penelitian didapatkan bahwa variabel usia hamil pertama ibu, jenis kelamin bayi, wilayah tempat tinggal, status bekerja ibu, sumber air minum berpengaruh signifikan terhadap kejadian BBLR di Kalimantan Selatan. Model klasifikasi yang dihasilkan memiliki nilai accuracy 0,844, sensitivity 0,852, spesificity 0,834, AUC 0,843 yang menunjukkan bahwa model cukup baik.
==============================================================================================================================
Infant Mortality Rate (IMR) is the number of deaths of children aged less than 12 months per 1000 live births. IMR is a public health indicator that reflects the living conditions of a country. In the context of IMR, the term 2/3 rule is known, which means that 2/3 of infant deaths occur during the neonatal period (0-28 days of age). Neonatal deaths are mostly caused by Low Birth Weight (LBW). South Kalimantan Province is a province with a very high IMR, namely 34 per 1000 live births in 2022. Apart from the IMR which is still high, South Kalimantan Province also has the highest percentage of LBW incidents in Indonesia, namely 6.8% and resulted in 180 neonatal deaths. This study aims to analyze the risk factors for LBW in South Kalimantan using a binary logistic regression model and the Synthetic Minority Oversampling Technique (SMOTE) approach to overcome unbalanced data proportions. The data used comes from the 2022 National Socio-Economic Survey (SUSENAS). The research results showed that the variables of mother's first pregnancy, baby's gender, area of residence, mother's working status, source of drinking water had a significant effect on the incidence of LBW in South Kalimantan. The resulting classification model has an accuracy value of 0.844, sensitivity 0.852, specificity 0.834, AUC 0.843 which shows that the model is quite good.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Angka Kematian Bayi, BBLR, Regresi Logistik Biner, SMOTE |
| Subjects: | H Social Sciences > HA Statistics > HA31.3 Regression. Correlation H Social Sciences > HA Statistics > HA31.7 Estimation H Social Sciences > HV Social pathology. Social and public welfare |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Ghevira Sandra E.Y |
| Date Deposited: | 08 Aug 2024 13:48 |
| Last Modified: | 27 Aug 2024 06:53 |
| URI: | http://repository.its.ac.id/id/eprint/115169 |
Actions (login required)
 |
View Item |