A Statistical Approach Towards Fraud Detection in the Horse Racing

With the inception of online betting in S. Korea, various foreigner professional gambling groups have exploited the betting regulations. This phenomenon has occurred mainly in Asia, because the regulations on gambling in these countries are complex and robust. Our study focuses on the horse racing in S. Korea, which is operated under the government funding. The foreigner gambling groups tried unlimited betting by modifying the official IoT (Internet of Things) based APP arbitrarily. We have checked that some abnormal transactions can occur by modifying this application. Our study proposes a fraud detection method that can help detecting abnormal activities and prevent them. Currently, the Korea Racing Authority (KRA) has been criticized for being ill-equipped to detect abnormal activities with the Walkerhill Incident. Our study presents a new anomaly detection model that uses a flexible window. In this study, we propose an idea that aims to detect abnormal betting transactions.