Abstract:

The primary goal of this research is to highlight how credit card evolution is a notable development in the banking sector. Every financial system has a sizable dataset for credit card transactions by customers. As a result, banks would require customer profiling as bank customers are aware of the issuer's determinations regarding who should receive banking facilities and what credit limit should be offered. Additionally, it aids issuers in developing a deeper understanding of their existing and potential clients. In earlier studies, customer profiling mostly relied on transaction data or demographic data; however, in the current study, both data are combined to produce a more accurate result and reduce risk. Finding the most effective method improves accuracy and aids banks in growing. By focusing on the important client (businesses), which are regarded as the key engine in the bank's profitability, finding the optimum technique improves accuracy and aids banks in achieving higher profitability. This study attempts to use fuzzy c means, improved k means, neural networks, and the k mean. The primary goal of this study is to develop a new label as a target for neural network classification using the labelled dataset. This will speed up clustering execution and produce the highest accuracy results. The accuracy ratio comparison demonstrates that the neural network is the best clustering technique.

Description:

.