EKSAKTA: Journal of Sciences and Data Analysis
VOLUME 3, ISSUE 2, August 2022

Comparison of the Naïve Bayes Classifier and Decision Tree J48 for Credit Classification of Bank Customers

Alifia Tanza (Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Indonesia)
Dina Tri Utari (Department of Statistics, Faculty of Mathematics and Natural Sciences, Universitas Islam Indonesia)



Article Info

Publish Date
29 Aug 2022

Abstract

The bank conducts an analysis or survey in the credit system to determine whether the customer is eligible to receive credit. With a case study of Bank BJB debtor data in December 2021, credit classification analysis was carried out by forming a model using the Naïve Bayes Classifier and Decision Tree J48. Thus it is expected to minimize the occurrence of bad loans. The data are divided into several categories: debtors with good, substandard, doubtful, and bad credit. The analysis was carried out using a 10-fold cross-validation model, where the results obtained from both tests, the highest accuracy value was the Decision Tree J48 of 78.26%. While the Naïve Bayes Classifier has a lower level of accuracy, the prediction results tend to be better than the Decision Tree J48. The prediction results with the Naïve Bayes Classifier can predict all classes and the most influential variable in classifying credit is the loan term.

Copyrights © 2022






Journal Info

Abbrev

eksakta

Publisher

Subject

Chemical Engineering, Chemistry & Bioengineering Chemistry Earth & Planetary Sciences Materials Science & Nanotechnology

Description

Ekstakta is an interdisciplinary journal with the scope of mathematics and natural sciences that is published by Fakultas MIPA Universitas Islam Indonesia. All submitted papers should describe original, innovatory research, and modelling research indicating their basic idea for potential ...