Behavioral Pattern Detection Analysis In Credit Cards Associated To Insurance Companies Using Deep Learning
Abstract
Fraud is any malicious action aimed at causing the other party to suffer financially. As digital money or plastic money is on the rise, even in developed countries, so is the fraud connected with it. Credit-card fraud has cost merchants and banks trillions of dollars worldwide. Even after various strategies for preventing fraud, fraudsters are actively finding new ways and techniques to commit fraud. The paper here deals to recommend an economical solution that automatically identifies insurance company-related swindle use credit cards using the deep learning algorithm, which is called as Autoencoders. Here the solution presented is based on autoencoder training to restore data, which is a standard form. Anomalies are generally observed through establishing a threshold for restoration error, and cases that have a higher threshold are considered anomalies.