Analisis Pemilihan Model Klasifikasi Penerima Subsidi Listrik Rumah Tangga Di Kota Surabaya Menggunakan Metode Cart, Naïve Bayes, Dan Regresi Logistik

Lestari, Tria Ayu (2019) Analisis Pemilihan Model Klasifikasi Penerima Subsidi Listrik Rumah Tangga Di Kota Surabaya Menggunakan Metode Cart, Naïve Bayes, Dan Regresi Logistik. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211540000060_Undergraduate_Thesis.pdf Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

Subsidi listrik merupakan salah satu upaya pemerintah Indonesia untuk memenuhi peningkatan kebutuhan energi listrik untuk masyarakat yang kurang mampu, sebagaimana yang diatur dalam UU Nomor 30 Tahun 2007 dan UU Nomor 30 Tahun 2009 tentang energi dan ketenagalistrikan. Namun dalam pelaksanaannya masih terdapat kendala yaitu berupa pemberian bantuan yang tidak tepat sasaran. Oleh karena itu, perlu untuk memastikan bahwa penerima manfaat subsidi listrik sudah tepat sasaran. Pada penelitian ini akan dilakukan pengklasifikasian terhadap status rumah tangga miskin penerima subsidi listrik daya 900 VA di Kota Surabaya yang dilakukan menggunkan metode Classification and Regression Tree (CART), Naïve Bayes Classifier, dan regresi logistik. Untuk mengatasi adanya imbalance class pada data penelitian, maka digunakan undersampling dan oversampling dengan 10 stratified fold validation. Sehingga dihasilkan kesimpulan bahwa dengan menggunakan metode regresi logistik dengan random undersampling yang menghasilkan nilai performa model yang lebih baik dibandingkan metode CART dan Naïve Bayes yaitu dengan nilai AUC, dan G-Mean yang lebih tinggi secara berurutan yaitu sebesar 59,75%, dan 60,89%.
=================================================================================================================================
The electricity subsidy is a form of concern from the Indonesian government to fulfill the increasing electricity needs of underprivileged communities, as regulated in Law Number 30 of 2007 and Law Number 30 of 2009 concerning energy and electricity. However, in its implementation, several problems have arisen, particularly because many beneficiaries are not part of the intended target group. Therefore, it is necessary to ensure that electricity subsidies are distributed to the appropriate recipients. In this study, the classification of the status of 900 VA electricity customers among poor households in Surabaya was carried out using the Classification and Regression Tree (CART) method, the Naïve Bayes Classifier, and logistic regression. The class imbalance in the research data was addressed using undersampling and oversampling techniques combined with 10-fold stratified cross-validation. The results indicate that the logistic regression method with random undersampling provides better model performance than both the Classification and Regression Tree (CART) method and the Naïve Bayes Classifier. This approach achieved the highest values of AUC and G-Mean, with the number
respectively are, 59,75%, and 60,89%.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	CART, Naïve Bayes Classifier, Regresi Logistik, Imbalance Class, Subsidi Listrik.
Subjects:	H Social Sciences > HA Statistics > HA29 Theory and method of social science statistics H Social Sciences > HD Industries. Land use. Labor > HD108 Classification (Theory. Method. Relation to other subjects ) H Social Sciences > HD Industries. Land use. Labor > HD7293 Housing policy
Divisions:	Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Tria Ayu Lestari
Date Deposited:	23 Jul 2026 02:44
Last Modified:	23 Jul 2026 02:44
URI:	http://repository.its.ac.id/id/eprint/67032

Actions (login required)

View Item