Perbandingan Algoritma SVM, Random Forest Dan XGBoost Untuk Penentuan Persetujuan Pengajuan Kredit
Mohammad Rizal Givari, Mochammad Riszky Sulaeman, Yuyun Umaidah
Abstract
Credit is an option for seeking funding for most economic activities. The demand for credit is currently growing very rapidly, in line with the increasing financial needs of the community, especially in developing countries such as Indonesia. Credit analysis needs to be carried out to achieve proper and safe lending. Credit analysis is an observation to see the feasibility of a credit problem. From this analysis, the creditworthiness of the recipient will be known. This study uses the CRISP-DM methodology which consists of 6 stages, namely Bussines Understanding, Data Understanding, Data preparation, Modeling Evaluation, and Deployment by applying the classification method by comparing the SVM, Random Forest, and XGBoost algorithms. This research uses an open source dataset obtained from Kaggle. The results of the research using the SVM, random forest, and XGBoost algorithms get the highest accuracy, recall, precision values in the XGBoost model with 82% accuracy, 70% recall, and 92% precision.