Prudent Fraud Detection Using Machine Learning

Abstract:
Most commercial Fraud Detection components of Internet banking systems use some kind of hybrid
setup usually comprising a Rule-Base and an Artificial Neural Network. Such rule bases have been
criticised for a lack of innovation in their approach to Knowledge Acquisition and maintenance.
Furthermore, the systems are brittle; they have no way of knowing when a previously unseen set of
fraud patterns is beyond their current knowledge. This limitation may have far reaching
consequences in an online banking system. This paper presents a viable alternative to brittleness in
Knowledge Based Systems; a potential milestone in the rapid detection of unique and novel fraud
patterns in Internet banking. The experiments conducted with real online banking transaction log
files suggest that Prudent based fraud detection may be a worthy alternative in online banking.

Existing system:
Traditionally, banks do not publicise specific details of their Fraud Detection (FD) systems . Despite this
fact, there have recently been a number of software vendors who publicised some information about
their Internet banking FD tools and the market share of such tools in commercial banks worldwide. For
example, the Proactive Risk Manager (PRM) was reported to be used by the top 20 banks in the world
and in more than 40 countries

Disadvantages:
 Results shows less accuracy compared to random forest classifier.

Proposed system:
This paper presents a viable alternative to brittleness in Knowledge Based Systems; a potential
milestone in the rapid detection of unique and novel fraud patterns in Internet banking. The
experiments conducted with real online banking transaction log files suggest that Prudent based
fraud detection may be a worthy alternative in online banking. We used randomforest classifier and
passive aggressive classifier algorithms to test the project accuracy.

Advantages:
The goal of this problem is to predict the status of the prudent fraud detection in banking transaction
of test dataset as accurate as possible.
Sensitivity, Specificity, Accuracy and it shows the error rate should be used to calculate the
performance of the proposed system process. The accuracy of logistic regression, Decision tree and
random forest classifier are 91.0, 92.3, and 96.5 respectively

Software Requirements:
Operating System : Windows

Language : Python

Framework : Django

HARDWARE REQUIREMENTS
RAM: 4GB and Higher
Processor : Intel i3 and above
Hard Disk: 500GB: Minimum