Riset Perancangan Model Prediksi Customer Churn dengan Graph Machine Learning PT. Bank Rakyat Indonesia

Meilani, Zahra Dyah (2023) Riset Perancangan Model Prediksi Customer Churn dengan Graph Machine Learning PT. Bank Rakyat Indonesia. Project Report. [s.n.], [s.l.].

Text 05111940000069-Project_Report.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (1MB) | Request a copy

Abstract

PT. Bank Rakyat Indonesia merupakan bank milik pemerintah terbesar di Indonesia. Salah satu produk dari BRI adalah BRImo yang merupakan aplikasi mobile banking untuk memudahkan nasabah BRI dalam melakukan transaksi perbankan. Pada setiap produk atau layanan, terdapat risiko pelanggan meninggalkan layanan atau produk tersebut karena berbagai alasan, baik dari internal maupun eksternal. Oleh karena itu, diperlukan model prediksi customer churn untuk mengetahui pengguna BRImo mana saja yang berisiko untuk meninggalkan layanan (churn) dan yang tidak. Model prediksi customer churn yang dikembangkan merupakan model machine learning berbasis graf dengan framework Topological Relational Inference – Graph Neural Network (TRI-GNN). Model yang dibuat mampu melakukan klasifikasi pengguna mana saja yang berpotensi untuk melakukan churn dengan memiliki nilai performa F1-score sebesar 0,752 dan recall 0,918.
===========================================================================================================================
PT. Bank Rakyat Indonesia is the largest government-owned bank in Indonesia. One of its products is BRImo, which is a mobile banking application designed to facilitate banking transactions for BRI customers. In every product or service, there exists a risk of customers leaving the service or product for various reasons, both internal and external. Therefore, a customer churn prediction model is required to identify which BRImo users are at risk of churning and which are not. The developed customer churn prediction model is a graph-based machine learning model utilizing the Topological Relational Inference – Graph Neural Network (TRI-GNN) framework. This model is capable of classifying potential churn users with a performance F1-score of 0.752 and a recall of 0.918.

Item Type:	Monograph (Project Report)
Uncontrolled Keywords:	BRIMO, Customer Churn, Graph Machine Learning, Topological Relational Inference – Graph Neural Network (TRI-GNN)
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA166 Graph theory
Divisions:	Faculty of Information Technology > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Zahra Dyah Meilani
Date Deposited:	31 Jul 2023 01:24
Last Modified:	31 Jul 2023 01:24
URI:	http://repository.its.ac.id/id/eprint/100735

Actions (login required)

View Item