Classification of Creditworthy Customer Using Support Vector Machine Algorithm

Abstract

The increase in the number of credit applications in the banking and financial institutions sector requires an efficient and accurate creditworthiness assessment system. Manually conducted assessments tend to be time-consuming and prone to subjectivity, so they can have an impact on errors in credit decision-making. The main problem faced is how to classify customers appropriately into creditworthy or non-creditworthy categories based on historical data. To overcome this, this study proposes the use of the Support Vector Machine (SVM) algorithm as an artificial intelligence-based solution that is able to handle classification problems with a high level of accuracy. The purpose of this study is to develop a model for classifying customer creditworthiness using the SVM algorithm and optimizing parameters to improve model performance. The methods used include the data preprocessing stage (handling missing values, categorical data encoding, and normalization), data division into training and test data, SVM model training, performance evaluation with accuracy, precision, recall, F1-score metrics, and parameter tuning using Grid Search and visualization through heatmaps. Kernel comparisons are also done to obtain the best configuration of the model. The results of the study show that the SVM model with the RBF kernel provides the best test performance reaching 87%, which means that the model is very good at recognizing potential customers who are not creditworthy. These results show that the SVM algorithm is effective in classifying customers' creditworthiness, so that it can be used as a decision-making tool in a more objective and efficient credit selection process.