<![CDATA[Credit is providing money or bills based on the agreement between a bank and another party. Lending is inseparable from bad credit risk, so credit analysis must be conducted on prospective debtors before approving a proposed loan. This research aims to analyze creditworthiness using a Classification Tree as a classification method with Random Oversampling to overcome imbalanced data. This study uses secondary data on the status of debtors from a bank in West Kalimantan. Research data amounted to 800 data samples consisting of collectability variables as target variables and 10 independent variables, namely limit, rate, tenor, total installments, age, salary, premium and admin, agency, type credit, and type need. The Classification Tree method with Random Oversampling is used to overcome imbalanced data. Classification begins with data preprocessing, then the data is divided into training and test data with proportions of 70:30, 80:20 and 90:10 for each treatment without Random Oversampling and with Random Oversampling. Next, a classification model is formed using training data, and the classification model is validated using test data. After that, an overall evaluation of the model is carried out to determine the best model used in the classification process. Based on the research results, the best model is the model Classification Tree with Random Oversampling in proportion 70:30, with an accuracy value of 89.17%, specificity of 75.00%, and recall of 89.66%. The model can be used to classify current and non-current debtor data. The most influential variable in classifying debtor status is the total installment variable.]]>
Copyrights © 2024
