Comparative analysis on vehicle insurances fraud detection using machine learning
H D Sheethal, Pabbisetty Pranavi, Sharanya S. Kumar, Sonika Kariappa, Baswaraju Swathi, Harinahalli Lokesh Gururaj · International journal of advance research, ideas and innovations in technology · 2020
Now-a-days frauds have become a serious threat to the society. Fraud is an illegal way of gaining more money. Frauds are posing problems to so many people. So, fraud detection becomes very important in this current world. Fraud detection can be implemented in various fields like banking, insurance, financial sectors and information security systems. And in the field of insurance, we have different types of insurances- health, vehicle and even life insurances. Frauds occur in each of these types of insurances. There are many approaches using which fraud can be detected. Machine learning, artificial intelligence, data mining and other methods are used to detect frauds. The well-known methods used in machine learning to detect frauds are Bayesian Network, Decision trees and back propagation techniques. Many algorithms are also used to detect frauds like Naive Bayes, KNN, Random forest. In this paper, the different techniques used for vehicle insurance fraud detection are presented along with comparative analysis.