A Differentiate Analysis for Credit Card Fraud Detection

With the swift progress of internet and electronic commerce, online money transaction has increased over time. People mostly eager to use online money transference and because of the internet is now available almost everywhere. Therefore, any attackers could be plan attacks from anywhere to forage any victim. There was various way from the previous attacks that victims became hunts by duplicate copy of the website, cards, ID numbers, rearranging provisional codes, fake documentations or signatures and so on. The genuine transaction and fraudulent transactions are almost similar, that's why it's very hard to figure out a real or fake transaction. One way could be effective if we know the behavioral pattern of card's owner. In this manner, we have introduced a Fraud Detection Engine (FDE) along with a Feature Selection Tool (FST). After gets a transaction request, FDE engine have searched for any sorts of intruder based on its key selective features. FST is one of those feature, which have used to match the cluster patterns with the requester behavioral pattern. Cluster pattern are cardholder's behavioral patterns which have trained by using of feedforward neural network. Therefore, examine all of the key features with vector analytical method or simply vector method. Proposed technique has applied from collected and previously driven on many studies datasets.