Prediksi Keberhasilan Program Pemasaran Listrik Prabayar Menggunakan Pembelajaran

Riduwan, Eko (2022) Prediksi Keberhasilan Program Pemasaran Listrik Prabayar Menggunakan Pembelajaran. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6032201194-Master_Thesis.pdf Restricted to Repository staff only Download (4MB)

Abstract

Listrik prabayar memberikan manfaat bagi perusahaan listrik dalam hal mengurangi piutang pelanggan dan bagi pelanggan, listrik prabayar memberikan kemudahan pengendalian pemakaian listrik. Oleh karena itu Perusahaan Listrik Negara (PLN) mempunyai program pemasaran untuk mengajak berpindah dari listrik pascabayar menjadi prabayar. Pencapaian KPI Program Pemasaran Prabayar PLN Tahuna 2021 hanya sebesar 52% dari taget 2.261 pelanggan beralih dari pascabayar ke prabayar. Hal ini memberikan peluang perbaikan mengingat program pemasaran saat ini belum mengoptimalkan penggunaan data sebagai dasar penentuan strategi pemasaran. Penelitian ini mengajukan metode identifikasi fitur dan skenario pemilihan algoritma Pembelajaran Mesin yang tepat untuk memprediksi penerimaan pelanggan listrik pascabayar terhadap program prabayar. Identifikasi fitur dilakukan dengan Spearman's rank correlation coefficient. Kandidat algoritma Pembelajaran Mesin yang terpilih adalah Logistic Regression, Support Vector Machines, Decision Tree, dan Random Forest. Model-model yang dihasilkan dievaluasi menggunakan confusion matrix sehingga didapatkan model terbaik untuk studi kasus yang diajukan. Penelitian menunjukkan bahwa fitur Tarif, Daya, frekuensi terlambat membayar listrik, pemakaian rata-rata listrik bulanan (kWh) dan Kabupaten mempunyai korelasi signifikan dan berpengaruh dengan penerimaan listrik prabayar. Adapun model dengan algoritma Random Forest adalah model terbaik sesuai tujuan penelitian dengan F-Measure tertinggi (95,173). Model prediksi terpilih selanjutnya di-deploy dalam bentuk Mobile App sehingga lebih mudah digunakan.
==============================================================================================================================
Prepaid electricity provides benefits for electricity companies in terms of reducing customer receivables, and for customers, prepaid electricity offers easy control of electricity consumption. Therefore, the State Electricity Company (PLN) has a marketing program to encourage switching from postpaid to prepaid electricity. The achievement of KPI for the 2021 PLN Prepaid Marketing Program is only 52% of the target of 2,261 customers switching from postpaid to prepaid. This KPI achievement provides an opportunity for improvement, considering that the current marketing program has not yet optimized the use of data as a basis for determining marketing strategies. This study proposes a feature identification method and scenario for selecting the suitable Machine Learning algorithm to predict postpaid electricity customer acceptance of prepaid programs. Spearman's rank correlation coefficient makes feature identification. The selected machine learning algorithm candidates are Logistic Regression, Support Vector Machines, Decision Tree, and Random Forest. The resulting models are evaluated using a confusion matrix to obtain the best model for the proposed case study. The research shows that the features of Tariff, Power, frequency of late payment of electricity, average monthly electricity usage (kWh), and Regency have a significant and influential correlation with prepaid electricity receipts. The model with the Random Forest algorithm is the best model according to the research objectives with the highest F-Measure (95.173). The selected prediction model is then deployed in the form of a Mobile App so that it is easier to use.

Item Type:	Thesis (Masters)
Additional Information:	RTMT 658.8 Rid p-1 2022
Uncontrolled Keywords:	Listrik prabayar, pemasaran, prediksi, pembelajaran mesin. Prepaid electricity, marketing, prediction, machine learning.
Subjects:	T Technology > T Technology (General)
Divisions:	Interdisciplinary School of Management and Technology (SIMT) > 61101-Master of Technology Management (MMT)
Depositing User:	Mr. Marsudiyana -
Date Deposited:	09 Jul 2026 03:31
Last Modified:	09 Jul 2026 03:31
URI:	http://repository.its.ac.id/id/eprint/134566

Actions (login required)

View Item