Analisis Bertahan Hidup Bayi Neonatal Menggunakan Pendekatan Multivariate Adaptive Regression Spline (MARS)

Putri, Atiqqah Edya (2025) Analisis Bertahan Hidup Bayi Neonatal Menggunakan Pendekatan Multivariate Adaptive Regression Spline (MARS). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5006211088-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB)

Abstract

Angka kematian neonatal, khususnya dalam 28 hari pertama kehidupan, menjadi indikator penting dalam menilai kualitas layanan kesehatan serta memiliki implikasi dalam bidang aktuaria, terutama dalam estimasi mortalitas dan perhitungan klaim asuransi. Penelitian bertujuan untuk menganalisis faktor yang memengaruhi kelangsungan hidup bayi neonatal di Kabupaten Donggala menggunakan metode Multivariate Adaptive Regression Splines (MARS). Data sekunder dari Dinas Kesehatan Kabupaten Donggala 2022–2024 digunakan sebanyak 101 data bayi, dengan 35% mengalami event dan 65% tersensor. Hasil analisis menunjukkan bahwa faktor seperti berat badan lahir rendah, jenis kelamin laki-laki, kehamilan pertama, usia kandungan prematur, usia ibu 24, 27, dan 30 tahun, tempat tinggal di daerah maju, cara persalinan spontan, serta penyakit atau kelainan meningkatkan risiko kematian neonatal. Bayi dari ibu berusia 23–25 tahun memiliki durasi bertahan hidup lebih lama. Model MARS menangkap hubungan nonlinier dan interaksi antar variabel prediktor, dengan 6 titik knot dan 5 basis fungsi yang terbentuk dalam model MARS menggunakan residual Cox-Snell dan 11 titik knot dan 6 basis fungsi yang terbentuk dalam model MARS menggunakan residual Cox-Snell. Model terbaik diperoleh menggunakan residual Cox-Snell dengan nilai GCV sebesar 0,124203 dan root mean squared error sebesar 0,314. Variabel signifikan dalam model adalah cara persalinan, penyakit/kelainan bayi, dan usia ibu kandung.
====================================================================================================================================
Neonatal mortality, particularly within the first 28 days of life, is a key indicator of healthcare quality and holds significant actuarial implications, especially in mortality estimation and health insurance claim calculations. This study aims to analyze the factors influencing neonatal survival in Donggala Regency using the Multivariate Adaptive Regression Splines (MARS) method. Using data from the Donggala District Health Office from 2022–2024, comprising 101 infants, with 35% event and 65% censored. The analysis revealed that low birth weight, male gender, first-time pregnancy, premature gestational age, maternal age 24, 27, and 30 years, residence in urban areas, spontaneous delivery, and the presence of congenital conditions or illnesses increase the risk of neonatal death. Additionally, infants born to mothers aged 23–25 years had a longer survival duration. MARS model capable of capturing nonlinear relationships and interactions among predictor variables, with 6 knots and 5 basis functions formed in the MARS model using Cox-Snell residuals, and 11 knots and 6 basis functions formed in the MARS model using Cox-Snell residuals. The best model was using Cox-Snell residuals, with GCV 0.124203 and a root mean squared error 0.314. The significant variables in the model are delivery method, neonatal disease/abnormality, and maternal age.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Analisis Survival, Bayi Neonatal, Kematian Bayi, MARS, Risiko Neonatal, Infant Death, MARS, Neonatal Infant, Neonatal Risk, Survival Analysis
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Divisions:	Faculty of Mathematics, Computation, and Data Science > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Atiqqah Edya Putri
Date Deposited:	31 Jul 2025 07:59
Last Modified:	31 Jul 2025 07:59
URI:	http://repository.its.ac.id/id/eprint/124334

Actions (login required)

View Item