Insurance Claim Fraud Detection Using Hidden Naive Bayes

Insurance claim fraud poses significant challenges for insurance companies, resulting in financial losses and a deterioration of industry trust. Reducing them requires correctly and successfully recognizing false insurance claims. This proposed method investigates the use of Hidden Naive Bayes in the identification of fraudulent insurance claims. An extension of Hidden Naive Bayes that increases the accuracy of fraud detection by utilizing hidden variables to capture latent variables, patterns in the Naive Bayes algorithm. This study is to investigate the potential applications of Hidden Naive Bayes in fraud detection and compare its efficacy to alternative fraud detection techniques. To find pertinent research papers, a thorough evaluation of the literature was done with an emphasis on the use of Hidden Naive Bayes in detecting insurance claim fraud. The proposed model recorded an accuracy of 78 %. The results reveal that by successfully modelling the links between observed and hidden factors, Hidden Naive Bayes shows potential in identifying phoney insurance claims. To improve the algorithm and investigate its sturdiness in managing massive and intricate insurance datasets, additional study is necessary. The work makes a contribution to the field of insurance fraud detection by emphasising the potential of Hidden Naive Bayes and offering details on its advantages, drawbacks, and prospective future research and application paths.