Pemodelan Faktor-Faktor yang Memengaruhi Kasus Pernikahan Dini pada Penduduk Generasi Z di Jawa Timur Menggunakan Rare Event Weighted Logistic Regression (RE-WLR)

Ardianata, Annas (2026) Pemodelan Faktor-Faktor yang Memengaruhi Kasus Pernikahan Dini pada Penduduk Generasi Z di Jawa Timur Menggunakan Rare Event Weighted Logistic Regression (RE-WLR). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003221033-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (5MB) | Request a copy

Abstract

Pernikahan dini masih menjadi permasalahan sosial yang dijumpai di berbagai daerah di Indonesia, termasuk pada penduduk Generasi Z di Provinsi Jawa Timur. Prevalensi pernikahan dini di Jawa Timur mencapai 7,78%, lebih tinggi dibandingkan rata-rata nasional dan tertinggi di antara provinsi-provinsi di Pulau Jawa. Kondisi ini menunjukkan adanya ketidakseimbangan data (imbalanced data) antara kelas menikah dini dan tidak menikah dini, sehingga diperlukan metode yang mampu menangani permasalahan tersebut secara metodologis. Penelitian ini bertujuan untuk mendeskripsikan karakteristik penduduk Generasi Z berdasarkan Status pernikahan dini serta mengidentifikasi faktor-faktor yang memengaruhinya menggunakan metode Rare Event Weighted Logistic Regression (RE-WLR). Data yang digunakan bersumber dari Survei Sosial Ekonomi Nasional (Susenas) Tahun 2024 di Provinsi Jawa Timur dengan unit analisis penduduk berusia 12 hingga 27 tahun. Metode RE-WLR diterapkan dengan menggabungkan teknik pembobotan, regularisasi ridge, dan koreksi bias untuk menangani ketidakseimbangan data sekaligus mengurangi overfitting. Proses pemodelan dilakukan menggunakan skema undersampling dengan tujuh rasio yang bervariasi dan sepuluh nilai parameter regularisasi (λ), sehingga diperoleh kombinasi optimal pada skema 1:30 dengan λ = 0,04. Model terbaik menghasilkan nilai AUC sebesar 78,80% yang termasuk kategori fair, dengan sensitivitas 79,44% dan spesifisitas 64,47%.. Hasil analisis menunjukkan bahwa tujuh variabel berpengaruh signifikan terhadap Status pernikahan dini, yaitu Tingkat pendidikan, Kepemilikan sumber penghasilan, Status bekerja, Kepemilikan jaminan kesehatan kategori Non-PBI, akses informasi televisi, status tempat tinggal, dan Kepemilikan fasilitas BAB. Temuan ini diharapkan dapat menjadi bahan pertimbangan bagi pemerintah dalam merumuskan kebijakan pencegahan pernikahan dini yang lebih tepat sasaran, khususnya di wilayah dengan prevalensi tinggi.
====================================================================================================================================
Early marriage remains a social problem found in various regions in Indonesia, including among Generation Z residents in East Java Province. The prevalence of early marriage in East Java reaches 7.78%, higher than the national average and the highest among provinces in Java. This condition indicates a data imbalance (imbalanced data) between the early marriage and non-early marriage classes, necessitating a method capable of addressing this issue methodologically. This study aims to describe the characteristics of Generation Z residents based on early marriage status and to identify the factors influencing it using the Rare Event Weighted Logistic Regression (RE-WLR) method. The data used were obtained from the National Socioeconomic Survey (Susenas) 2024 in East Java Province, with the unit of analysis being residents aged 12 to 27 years. The RE-WLR method was applied by combining weighting techniques, ridge regularization, and bias correction to handle data imbalance while reducing overfitting. The modeling process was carried out using an undersampling scheme with seven varying ratios and ten regularization parameter values (λ), yielding an optimal combination at a 1:30 scheme with λ = 0.04. The best model produced an AUC value of 78.80%, classified as fair, with a sensitivity of 79.44% and specificity of 64.47%. The analysis results indicate that seven variables significantly influence early marriage status, namely education level, ownership of income sources, employment status, ownership of Non-PBI health insurance, television information access, place of residence, and ownership of sanitation facilities. These findings are expected to serve as considerations for the government in formulating more targeted early marriage prevention policies, particularly in areas with high prevalence.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Ketidakseimbangan Data, Pernikahan Dini, RE-WLR, Regularisasi, Undersampling. Early Marriage, Imbalanced Data, RE-WLR, Regularization, Undersampling
Subjects:	H Social Sciences > HA Statistics H Social Sciences > HA Statistics > HA31.3 Regression. Correlation. Logistic regression analysis. H Social Sciences > HA Statistics > HA31.7 Estimation Q Science Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Annas Ardianata
Date Deposited:	30 Jul 2026 02:22
Last Modified:	30 Jul 2026 02:22
URI:	http://repository.its.ac.id/id/eprint/140092

Actions (login required)

View Item