Pemanfataan Machine Learning Untuk Prediksi Laporan Gangguan Pelanggan (Studi Kasus : PLN UP3 Makassar Selatan)

Budiyono, Yanuardhi Arief (2022) Pemanfataan Machine Learning Untuk Prediksi Laporan Gangguan Pelanggan (Studi Kasus : PLN UP3 Makassar Selatan). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6032201102-MASTER_THESIS.pdf Restricted to Repository staff only Download (2MB)

Abstract

PT PLN (Persero) terus berupaya meningkatkan layanan kepada pelanggan dalam hal durasi penanganan laporan gangguan pelanggan melalui implementasi command center. Kompleksitas data laporan gangguan pelanggan yang tersimpan pada database command center belum dilakukan analisis untuk memperkirakan jumlah laporan dan durasi penanganan gangguan pelanggan dimasa yang akan datang. Penelitian ini bertujuan untuk memprediksi jumlah laporan dan durasi penanganan gangguan pelanggan agar mengetahui apakah penanganan gangguan memenuhi target kinerja atau tidak. Metode penelitian menggunakan regresi linier berganda untuk prediksi jumlah laporan gangguan dan durasi penanganan gangguan. Perbandingan metode klasifikasi decision tree dan naïve bayes dilakukan untuk prediksi terpenuhi atau tidak terpenuhi durasi penanganan gangguan. Hasil penelitian menunjukkan nilai R2 terbaik prediksi jumlah laporan 0,698 dan 0,566 untuk durasi penanganan gangguan. Nilai RMSE terbaik prediksi jumlah laporan 12,067 dan 0,31 untuk durasi penanganan gangguan. Hasil klasifikasi didapatkan ROC curve terbaik sebesar 0,979 pada metode naïve bayes. Mengacu nilai akurasi, model dapat diimplementasikan pada sistem command center untuk mengatur komposisi dan jumlah tim pelayanan teknik serta sebagai usulan metode kontrak variable cost pelayanan teknik.
==============================================================================================================================
PT PLN (Persero) continues to improve services to handling the recovery time of customer complaints through the command center. The complexity of customer complaints data stored in the command center database has not been analyzed to estimate the number of complaints and the recovery time in the future. This study aims to predict the number of complaints and the recovery time in order to find out whether the recovery time fullfiling the performance targets or not. The research method uses multiple linear regression to predict the number of customer complaints and the recovery time. Comparison of the decision tree and naïve bayes classification methods was carried out to predict target of recovery time fulfilled. The results showed that the best predicted R2 value for customer complaints around 0.698 and 0.566 for the reovery time. The best RMSE values of customer complaints around 12.067 and 0.31 for the recovery time. The classification results obtained the best ROC curve of 0.979 in the naïve bayes method. Referring to the accuracy value, the model can be implemented in the command center system to regulate the composition and number of the pelayanan teknik team as well as a proposed method of contracting variable cost pelayanan teknik.

Item Type:	Thesis (Masters)
Additional Information:	RTMT 658.834 2 Bud p-1 2022
Uncontrolled Keywords:	Laporan Gangguan Pelanggan, Machine Learning, Regresi Linier Berganda, Klasifikasi. Customer Complaint Reports, Machine Learning, Multiple Linier Regression, Decision Tree.
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:	08 Jul 2026 02:30
Last Modified:	08 Jul 2026 02:30
URI:	http://repository.its.ac.id/id/eprint/134486

Actions (login required)

View Item