NH Nonghyup Card Completes FDS Advancement Project

The Card Industry's First '2 Stage Model Operation Process' Established

[Asia Economy Reporter Ki Ha-young] NH Nonghyup Card announced on the 20th that it has completed the advancement project of the Fraud Detection System (FDS) to protect customers from increasingly sophisticated and large-scale financial fraud.

Through this advancement, NH Nonghyup Card introduced the industry's first '2-Stage Model Operation Process' system utilizing multiple machine learning algorithms. To improve the domestic credit card fraud prediction model and detect unusual patterns, they newly established ▲Self-Anomaly Transaction Detection Model ▲Card Voice Phishing Score Model development and monitoring environment.

This system is a process that separates the card fraud prediction model into two types: score model and sub-model. Specifically, the score model quantifies the possibility of abnormal transactions by analyzing past fraud patterns to primarily prevent financial fraud, and secondarily, sub-models detect and supplement it.

Additionally, they introduced functions that automatically generate rule-based methods essential for detecting abnormal transactions and added unusual pattern analysis functions using machine learning technology.

An NH Nonghyup Card official said, "Our goal is to build a data environment that can protect customers by identifying not only fraud areas but also incidents caused by unusual transactions in advance," adding, "We will continue to lead improvements in FDS performance by including areas not attempted by domestic card companies in the management system and strive to establish NH Nonghyup Card as a safe card company."

This content was produced with the assistance of AI translation services.