Optimized Deep Learning for Car Insurance Fraud Detection Using Gray Wolf Algorithm

As policy knowledge grows, customers are now able to take advantage of a wide range of benefits, including property insurance, travel insurance, health insurance, and auto insurance. With more people opting to buy such insurance, con artists can easily take advantage of them. Clients and vendors of insurance contracts may commit insurance fraud. Client-side insurance fraud includes overestimated claims and postdated policies. In the vendor side of insurance fraud, policies are issued by non-existent companies, and premiums are not submitted, among other things. Deep learning can also help prevent fraudsters by spotting abnormal trends in insurance claims processing and customer background checks, possibly saving insurers a lot of money. This paper proposed Car Insurance Fraud Detection (CIFD) with hyper parameter optimization. The Gray Wolf optimization model shows optimal predictions of car insurance fraud using deep learning. Performance metrics such as accuracy are used to assess the effectiveness of proposed algorithms. CIFD has the highest accuracy rate of 99% among all deep learning models.