Analisis Faktor yang Memengaruhi Status Desa di Provinsi Papua Menggunakan Metode Regresi Probit dengan Pendekatan Combine Sampling

Prastyo, Yoppy Eka (2024) Analisis Faktor yang Memengaruhi Status Desa di Provinsi Papua Menggunakan Metode Regresi Probit dengan Pendekatan Combine Sampling. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 5003201097-Undergraduate_Thesis.pdf] Text
5003201097-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (2MB) | Request a copy

Abstract

Pembangunan nasional di Indonesia, terutama di wilayah pedesaan, menjadi fokus utama untuk mencapai tujuan pembangunan berkelanjutan. Meskipun berbagai upaya pembangunan telah dilakukan, terdapat ketidakmerataan distribusi pembangunan di seluruh daerah, khususnya di Provinsi Papua yang menghadapi tantangan tingkat ketertinggalan yang signifikan. Penelitian ini mengambil pendekatan klasifikasi dengan menggunakan Regresi Probit dan metode Combine Sampling (gabungan SMOTE dan Tomek Links) untuk mengatasi ketidakseimbangan data pada status desa tertinggal di Provinsi Papua tahun 2021. Analisis statistik deskriptif menunjukkan bahwa 92,9% desa di Papua tergolong tertinggal, sementara 7,1% tidak tertinggal. Dari 12 variabel prediktor yang dianalisis, sebagian besar desa memiliki nilai minimum 0 untuk variabel seperti SD/MI, keluarga pengguna listrik, unit usaha BUMDes, pasar, kelompok pertokoan, dan bank umum serta BPR. Beberapa desa juga mengalami pencemaran air dan minim sistem peringatan dini bencana alam, menunjukkan ketimpangan dalam aspek sosial, ekonomi, dan lingkungan. Hasil resampling menggunakan metode combine sampling menunjukkan peningkatan keseimbangan data antara desa tertinggal dan tidak tertinggal. Analisis regresi probit dengan data resampling mengidentifikasi 10 variabel signifikan yang memengaruhi status desa, seperti jumlah SD/MI, rasio keluarga pengguna listrik, dan kejadian pencemaran air. Model regresi probit yang dihasilkan menunjukkan akurasi yang baik dengan nilai rata-rata AUC 86,1%, G-mean 80,8%, akurasi klasifikasi 80,8%, sensitivitas 79,2%, dan spesifisitas 82,6%. Fold terbaik menunjukkan nilai AUC 87,87% dan akurasi klasifikasi 81,72%. Hasil penelitian ini diharapkan dapat membantu pemerintah merancang kebijakan pembangunan yang lebih akurat dan tepat sasaran di Provinsi Papua.
============================================================
National development in Indonesia, especially in rural areas, is the main focus to achieve sustainable development goals. Despite various development efforts, there is an uneven distribution of development across regions, particularly in Papua Province which faces significant underdevelopment challenges. This study takes a classification approach using Probit Regression and Combine Sampling technique (combined SMOTE and Tomek Links) to overcome data imbalance in the status of underdeveloped villages in Papua Province in 2021. Descriptive statistical analysis shows that 92.9% of villages in Papua are classified as underdeveloped, while 7.1% are not underdeveloped. Of the 12 predictor variables analyzed, most villages have a minimum value of 0 for variables such as primary school/middle school, families using electricity, BUMDes business units, markets, groups of shops, and commercial banks and BPRs. Some villages also experienced water pollution and lacked early warning systems for natural disasters, indicating inequality in social, economic, and environmental aspects. Resampling results using the combined sampling method showed an improvement in data balance between disadvantaged and non-disadvantaged villages. Probit regression analysis with the resampled data identified 10 significant variables affecting village status, such as the number of primary schools, the ratio of families using electricity, and the incidence of water pollution. The resulting probit regression model showed good accuracy with an average AUC value of 86.1%, G-mean of 80.8%, classification accuracy of 80.8%, sensitivity of 79.2%, and specificity of 82.6%. The best fold showed an AUC value of 87.87% and a classification accuracy of 81.72%. The results of this study are expected to help the government design more accurate and targeted development policies in Papua Province.

Item Type: Thesis (Other)
Uncontrolled Keywords: Combine Sampling, Desa Tertinggal, Provinsi Papua, Regresi Probit, Status Desa. Backward Villages, Combine Sampling, Papua Province, Probit Regression, Village Status.
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: Yoppy Eka Prastyo
Date Deposited: 09 Aug 2024 04:10
Last Modified: 09 Aug 2024 04:10
URI: http://repository.its.ac.id/id/eprint/115073

Actions (login required)

View Item View Item