<![CDATA[The banking world in terms of providing credit to customers is a regular activity that has a large effect. In its application, non-performing loans or bad loans are often created due to poor credit analysis in the credit granting process, or from bad customers. The purpose of this study is to compare the results of algorithm accuracy between K-Nearest Neighbor (K-NN), Decision Tree, and Naive Bayes which results in the best accuracy will be implemented to determine creditworthiness. The attributes used in this study consisted of 11 attributes, namely marital status, number of dependents, age, last education, occupation, monthly income, home ownership, collateral, loan amount, length of loan and information as result attributes. The methods used in this research are K-Nearest Neighbor, Decision Tree, and Naive Bayes. From the results of evaluation and validation using k-5 fold that has been carried out using RapidMiner tools, the highest accuracy results from a comparison of 3 algorithms is using a decision tree (C4.5) of 98% in the 3rd test.]]>
Copyrights © 2021
