“AI also explains how to raise credit rating”… Research paper published by Shinhan Bank

(Provided by Shinhan Bank) © News1

Shinhan Bank is taking a step closer to developing an algorithm that can evaluate customers’ creditworthiness through AI (artificial intelligence) and even explain how to raise creditworthiness.

Shinhan Bank announced on the 9th that the paper of’A Study on Credit Rating Model Using Explainable AI Technology’ by AICC (Integrated AI Center) in the Digital Innovation Group has been approved for publication by the Korean Journal of Data Information Science (KCI’s excellent journal).

AI technology is difficult to understand and explain the process to the extent that it compares the process of obtaining results to a’black box’. Shinhan Bank’s AICC conducted this study based on the judgment that applying AI to the financial industry requires not only high precision but also the ability to explain the results.

In this paper, AICC of Shinhan Bank proposed an algorithm that can explain the effect of each item of credit rating on the result value, along with an AI-based credit rating algorithm, and the reason for fluctuations in the result value derived by AI.

Through this study, it was possible to explain to customers the impact of individual items, such as salary level and debt amount, among customers’ credit rating components during AI credit evaluation. Even when a customer’s credit rating fluctuates, it can intuitively explain which item the credit rating has risen or declined.

In addition, Shinhan Bank AICC conducted a study to improve the explanatory power of financial AI models with Korea University’s pattern recognition and machine learning labs. As a result, AI can provide alternatives by analyzing credit evaluation items such as salary, delinquency count, utility bill arrears, and installment transaction weight for customers who refused loans.

A Shinhan Bank official said, “This study is of great significance in that it has confirmed that it can provide explanatory power when introducing AI to financial services.” “Through continuous XAI (Explainable AI) research, customer-oriented innovative digital services I will show you.”

[email protected]