Article Per Year (5 Year)
p-Index From 2020 - 2025
0.23
P-Index
This Author published in this journals
All Journal Jurnal Algoritma
Ananda Surya, Archie
Unknown Affiliation
Computer Science & IT

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Prediksi Kapabilitas Calon Debitur Menggunakan Analisis Data Machine Learning Dengan Metode Random Forest Ananda Surya, Archie; Rizki Darmawan, Dika; Solichin, Achmad
Jurnal Algoritma Vol 22 No 1 (2025): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.22-1.1929

Abstract

This study was conducted at a financing company engaged in investment, working capital, and multipurpose financing, particularly in the consumer financing sector for four-wheeled vehicles, two-wheeled vehicles, and multipurpose products. The high level of non-performing loans (NPLs) has led to a decline in the company's asset quality. Although credit scoring has been implemented as the initial filter to evaluate prospective borrowers, certain aspects are still being overlooked. Credit scoring becomes less relevant when a prospective borrower has little or no credit history at all—a situation commonly encountered in motorcycle or two-wheeled vehicle financing. Therefore, this research aims not only to provide insights into borrower classification but also to reduce the percentage of non-performing loans, especially in two-wheeled vehicle financing in the future. The classification method used is Random Forest with SMOTE, utilizing 28 parameters including credit scoring, salary, number of active loans, borrower age, application region, marital status, length of employment, gender, and number of dependents. The data used for training the model in this study consists of records available in the system up to the year 2023, totaling 254,609 entries. The research findings indicate that the Random Forest model achieved an accuracy rate of 69%, which is slightly higher than the LightGBM (LGBM) method at 59% and XGBoost, which had an accuracy rate of 39%.
Co-Authors Achmad Solichin Rizki Darmawan, Dika