Towards the Improvement of Credit Card Approval Process Using Classification Algorithm

For financial institutions, such as banks, one of the major sources to earn includes the issuance of credit cards. Over time, there has been a significant growth of applicants that opt to take advantage of the convenience and staggered payments provided by credit cards. As such, financial institutions are faced with the tedious and daunting task of reviewing and approving applications on a periodic basis. In order to handle fatigue-induced errors as well as increase reliability and efficiency in the review and approval process, financial institutions employ digitalization as a support system mainly in the form of machine learning algorithms. Among the set of alternative algorithms available for use, financial institutions are still prompted with the task of evaluating these algorithms based on performance metrics Such a way tends to ensure that the credit card approval process can be improved further. Therefore, this paper uses a classification algorithm run through WEKA to explore the different algorithms' performance and their impact on the process improvement of credit card applications. It is found that financial institutions consider the following attributes important in the process of approval: default history, credit score status, income status, employment history, and banking history. Under such attributes, the algorithm was also able to present a sound and accurate classification task.