Pemodelan Faktor-Faktor yang Memengaruhi Kasus KDRT di Provinsi Jawa Timur Menggunakan Regresi Logistik Biner Tanpa dan Dengan SMOTE-NC (Studi Semi Sensus Keluarga 2021)

Agustin, Putri Emilia (2024) Pemodelan Faktor-Faktor yang Memengaruhi Kasus KDRT di Provinsi Jawa Timur Menggunakan Regresi Logistik Biner Tanpa dan Dengan SMOTE-NC (Studi Semi Sensus Keluarga 2021). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003201051-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Kekerasan dalam Rumah Tangga (KDRT) merupakan konflik yang berdampak paling parah karena melibatkan kekerasan fisik, seksual, psikologis, dan penelantaran rumah tangga. Provinsi Jawa Timur menjadi salah satu provinsi dengan kasus kekerasan terbanyak dan mayoritas korbannya yaitu perempuan. Meskipun kasus KDRT cukup sedikit, tetapi berdampak besar bagi ketahanan keluarga sehingga diharapkan tidak terjadi. Oleh karena itu, dilakukan pemodelan dan identifikasi faktor-faktor yang memengaruhi kasus KDRT pada perempuan berstatus kawin di Provinsi Jawa Timur menggunakan metode regresi logistik biner dengan pendekatan SMOTE-NC (Synthetic Minority Oversampling Technique - Nominal Continuous). Data yang digunakan bersumber dari Pendataan Keluarga tahun 2021 oleh BKKBN dengan unit penelitian perempuan berstatus kawin di Provinsi Jawa Timur. Terdapat 5.582.645 perempuan yang tidak mengalami KDRT (99,66%) dan 19.092 perempuan mengalami KDRT (0,34%). Mayoritas perempuan mempunyai usia kawin pertama 20–25 tahun. Perempuan yang mengalami KDRT mayoritas berpendidikan akhir tamat SD/Sederajat dengan mayoritas jumlah anak masih hidup 1–2 anak. Pemodelan regresi logistik biner menghasilkan nilai sensitivitas 0%, spesifisitas 100%, G-mean 0%, dan AUC 50% dengan 11 variabel prediktor yang berpengaruh signifikan terhadap kasus KDRT. Model regresi logistik biner dengan penanganan SMOTE-NC yang terbaik menggunakan proporsi kategori KDRT sebesar 20%. Nilai sensitivitas 10,61%, spesifisitas 98,37%, G-mean 32,30%, dan AUC 54,49%. Terdapat 11 variabel prediktor berpengaruh signifikan terhadap kasus KDRT di Provinsi Jawa Timur, yaitu usia kawin pertama, tingkat pendidikan, kepemilikan anak, interaksi antar anggota keluarga, jenis pengasuhan anak, rekreasi bersama keluarga, keikutsertaan kegiatan sosial, terdapat anggota keluarga yang sakit, kepemilikan sumber penghasilan, kepemilikan tabungan, dan kepemilikan usaha ekonomi.
=======================================================================================================
Domestic Violence is a conflict that has the most severe impact because it involves physical, sexual, psychological violence and domestic neglect. East Java Province is one of the provinces with the most cases of violence and the majority of victims are women. Even though there are quite a few cases of domestic violence, it has a big impact on family resilience so it is hoped that this will not happen. Therefore, modeling and identification of factors influencing domestic violence cases among married women in East Java Province were carried out using the binary logistic regression method with the SMOTE-NC (Synthetic Minority Oversampling Technique - Nominal Continuous) approach. The data used comes from the 2021 Family Data Collection by the BKKBN with a research unit for married women in East Java Province. There were 5,582,645 women who did not experience domestic violence (99.66%) and 19,092 women experienced domestic violence (0.34%). The majority of women have a first marriage age of 20–25 years. The majority of women who experience domestic violence have primary school/equivalent education with the majority having 1-2 living children. Binary logistic regression modeling produced a sensitivity value of 0%, specificity 100%, G-mean 0%, and AUC 50% with 11 predictor variables that had a significant effect on cases of domestic violence. The binary logistic regression model with the best SMOTE-NC treatment uses a proportion of the domestic violence category of 20%. The sensitivity value was 10.61%, specificity 98.37%, G-mean 32.30%, and AUC 54.49%. There are 11 predictor variables that have a significant influence on domestic violence cases in East Java Province, namely age at first marriage, level of education, ownership of children, interactions between family members, type of child care, recreation with family, participation in social activities, presence of sick family members, ownership of resources. income, savings ownership, and economic business ownership.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Imbalanced Data, Jawa Timur, KDRT, Regresi Logistik Biner, SMOTE-NC, Binary Logistic Regression, Domestic Violence, East Java, Imbalanced Data, SMOTE-NC
Subjects:	H Social Sciences > HA Statistics > HA31.3 Regression. Correlation H Social Sciences > HN Social history and conditions. Social problems. Social reform H Social Sciences > HQ The family. Marriage. Woman
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Putri Emilia Agustin
Date Deposited:	09 Aug 2024 04:00
Last Modified:	09 Aug 2024 04:00
URI:	http://repository.its.ac.id/id/eprint/114945

Actions (login required)

View Item