Pemodelan Status Rumah Tangga Miskin di Daerah Istimewa Yogyakarta Menggunakan Regresi Probit Ordinal SMOTE-NC

Novimbriani, Dita (2025) Pemodelan Status Rumah Tangga Miskin di Daerah Istimewa Yogyakarta Menggunakan Regresi Probit Ordinal SMOTE-NC. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211035-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Kemiskinan merupakan permasalahan kompleks dan multidimensional yang mencakup aspek sosial, ekonomi, budaya, dan lainnya. Daerah Istimewa Yogyakarta (DIY) menempati peringkat tertinggi tingkat kemiskinan di Pulau Jawa dari tahun 2019 hingga 2024, dengan persentase penduduk miskin sebesar 10,83% pada Maret 2024. Identifikasi faktor-faktor penyebab kemiskinan di berbagai tingkatan rumah tangga sangat penting untuk memahami kondisi kemiskinan di setiap tingkat rumah tangga. Penelitian ini bertujuan untuk melakukan pemodelan dan identifikasi faktor-faktor yang mempengaruhi status rumah tangga miskin di DIY dengan pendekatan Synthetic Minority Oversampling Technique-Nominal Continuous (SMOTE-NC). Data yang digunakan bersumber dari Badan Pusat Statistik (BPS) dengan unit penelitian rumah tangga. Status rumah tangga miskin dikategorikan menjadi 3 yaitu tidak miskin, rentan miskin dan miskin. Terdapat 67,3% rumah tangga tidak miskin, 23,8% rumah tangga rentan miskin dan 9% rumah tangga miskin. Terlihat bahwa terjadi ketidakseimbangan data. Karakteristik rumah tangga miskin mayoritas memiliki jumlah anggota rumah tangga lebih banyak, kepala rumah tangga berjenis kelamin perempuan, berusia lebih tua, berstatus kawin, berpendidikan rendah (SD atau lebih rendah), bekerja di sektor pertanian, dan tinggal di wilayah perdesaan. Model regresi probit ordinal tanpa SMOTE-NC menghasilkan ketepatan klasifikasi 69%, presisi 53%, sensitivitas 42%, dan spesifisitas 70% dengan lima variabel prediktor yang signifikan. Setelah dilakukan penyeimbangan data menggunakan SMOTE-NC, diperoleh ketepatan klasifikasi 60%, presisi 50%, sensitivitas 46%, dan spesifisitas 73% dan terdapat enam variabel prediktor yang signifikan yaitu. jumlah anggota rumah tangga, umur Kepala rumah tangga, status perkawinan Kepala rumah tangga, pendidikan tertinggi Kepala rumah tangga, lapangan pekerjaan Kepala rumah tangga dan klasifikasi wilayah tempat tinggal.
======================================================================================================================================
Poverty is a complex and multidimensional problem that includes social, economic, cultural, and other aspects. The Special Region of Yogyakarta (DIY) ranked the highest poverty rate on the island of Java from 2019 to 2024, with a percentage of the poor population of 10.83% in March 2024. Identification of the factors that cause poverty at various levels of households is very important to understand the conditions of poverty at each level of households. This study aims to model and identify the factors that affect the status of poor households in DIY with Synthetic Minority Oversampling Technique-Nominal Continuous (SMOTE-NC) approach. The Data used was sourced from the Central Statistics Agency (BPS) with the household research unit. The Status of poor households is categorized into 3, namely non-poor, vulnerable poor and poor.There were 67,3% non poor households, 23,8% vulnerable households and 9% poor households. It appears that there is an imbalance of data. Characteristics the majority of poor households have more household members, female head of household, older age, married status, low education (elementary school or lower), work in the agricultural sector, and live in rural areas. Ordinal probit regression Model without SMOTE-NC resulted in 69% classification accuracy, 53% precision, 42% sensitivity, and 70% specificity with five significant predictor variables. After balancing the data using SMOTE-NC, obtained classification accuracy of 60%, 50% precision, sensitivity of 46%, and specificity of 73% and there are six significant predictor variables, namely. the number of household members, the age of the head of the household, the marital status of the head of the household, the highest education of the head of the household, the occupation of the head of the household and the classification of the region of residence.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Kemiskinan, Regresi, Probit, Ordinal, SMOTE-NC, Poverty, Regression,Probit, Ordinal, SMOTE-NC
Subjects:	H Social Sciences > HA Statistics > HA29 Theory and method of social science statistics
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Dita Novimbriani
Date Deposited:	29 Jul 2025 03:09
Last Modified:	29 Jul 2025 03:09
URI:	http://repository.its.ac.id/id/eprint/122478

Actions (login required)

View Item