Description
ABSTRACT
In the present time, the internet has become more popular in almost every domain of life. Online shopping, e- commerce and transactions are increasing every day. Fraudulent activities have also increased in the online payment system. Payment card fraud has become a serious problem in the world. Companies and institutions lose huge amounts annually due to fraud and the fraudsters continuously seek new ways to commit illegal actions. The good news is that fraud tends to be perpetrated to certain patterns and that it is possible to detect such patterns, and hence fraud. In this paper, we present a way to detect fraudulent transactions by a neural network model. Artificial neural network (ANN), when trained properly can work like a human brain. They learn by example, like people and are known as a very good classifier. Among the main characteristics of credit card traffic are the great imbalance between proper and fraudulent operations. At the same time, public data are hardly available for confidentiality issues. To deal with the imbalanced dataset we use resampling techniques. To ensure proper model construction we made a pattern recognition network trained with a scaled conjugate gradient back propagation algorithm using the Neural network.
TABLE OF CONTENTS
COVER PAGE
TITLE PAGE
APPROVAL PAGE
DEDICATION
ACKNOWLEDGEMENT
ABSTRACT
CHAPTER ONE
- INTRODUCTION
- BACKGROUND OF THE PROJECT
- PROBLEM STATEMENT
- SCOPE OF THE PROJECT
- SIGNIFICANCE OF THE STUDY
- PURPOSE OF THE PROJECT
- AIM/OBJECTIVE OF THE PROJECT
CHAPTER TWO
LITERATURE REVIEW
- DATA ANALYSIS TECHNIQUES FOR FRAUD DETECTION
- OVERVIEW OF PATTERN RECOGNITION
- FRAUD DETECTION AND PREVENTION
- CREDIT CARD FRAUD TYPES
- THE DETECTION OF CREDIT CARD FRAUD
- REVIEW OF CREDIT CARD FRAUD DETECTION METHODS
- OVERVIEW OF CREDIT CARD FRAUD
CHAPTER THREE
METHODOLOGY
- CREDIT CARD FRAUD AND NEURAL NETWORKS
- HANDLING IMBALANCED DATASETS
- PREPARING THE DATASET
CHAPTER FOUR
- BUILDING THE MODEL
CHAPTER FIVE
- CONCLUSION
- RECOMMENDATION
- REFERENCES
- CHAPTER ONE
- 1.0 INTRODUCTION
The concept of fraud is present in the earliest writings of history and has since developed into an evolutionary subset of financialfraud (William Brock and Marc-AndréBoutin, 2012). UK legislation defines fraud as “An Act to make provision for, and in connection with, criminal liability for fraud and obtaining services dishonestly.” This is found in the legally binding UK Legislation 2006 that states the criminal proceedings for the specific actions undertaken by fraudsters. Fraud today comprises of many different types, such as pyramid schemes, identity theft, and credit card fraud. This project focuses on the constant evolution of credit card fraud techniques and how it has influenced the analysis of the different credit card transaction processes that fraudsters take advantage of.
Nowadays more and more people prefer to use credit cards when purchasing goods and services [1]. Credit cards are very useful when shopping online as well as when paying bills and taxes online. Not only that they are convenient, but they also save time. Many people find credit card payment in stores much more practical than cache payments. However, with the increase in the number of credit card users, the number of credit card frauds also increases. Neural network applications have great potential to detect and prevent fraud.
The goal of credit card fraud detection is to decide whether a transaction is fraudulent based on historical data. It is not easy to decide because of changes in client spending practices, especially during the holiday season. Fraudsters use different techniques to overcome fraud protection. It is currently realized that artificial neural networks offer a successful way to deal with problems like these [2].
According to Pozzolo et al. [3], one of the most explored fields of fraud detection [4, 5, 6] is credit card fraud detection. It detects fraudulent behavior by automated analysis of previous transaction data. Many attributes (e.g. recipient, quantity of the transaction, date of transaction, credit card identifier) are stored in the databases for every transaction. But fraud occurrence could be detected only of the credit card level because one transaction is not sufficient [4]. All transactions for one credit card are grouped for learning behavioral models. The data for each card is analyzed to recognize fake transactions [7].
Because there is a large volume of customer data and transactional data, neural networks can be effectively used for the detection of credit card behavior and usage patterns. Neural network applications can detect and prevent fraud by analyzing the irregularities in the pattern changes.
1.1 BACKGROUND OF THE PROJECT
The growing development of ecommerce has made payment cards the most popular payment method for online purchases. This combined with the advancement of fraudulent attacks has created a need to detect fraudulent transactions with advanced fraud detection algorithms (S. Benson Edwin Raj and A. Annie Portia, 2011). As the nature of credit card fraud is ever-changing, industries need to detect credit card fraud using fraud detection algorithms (FDAs) given the current threat landscape.
1.2 PROBLEM STATEMENT
Use of credit and debit cards has increased vigorously in the last years, unfortunately so has the fraud committed with them. In Nigeria, between 2010 and 2019 the total level of fraud committed via credit card highly increased. Hence it is still a challenge to find effective methods to mitigate fraud. Existing solutions include simple if-then rules and classical machine learning algorithms. From an academic perspective, credit card fraud detection is a standard classification problem, in which historical transaction data is used to predict future frauds. However, practical aspects make the problem more complex. Indeed, existent comparison measures lack a realistic representation of monetary gains and losses, which is necessary for effective fraud detection. Moreover, there is an enormous amount of transactions from which only a tiny part are frauds, which implies a huge class imbalance often dealt with by stratified sampling leading to data samples with incomplete information. The large increase in online credit card transactions is another issue that calls for tailored solutions. Finally, a real fraud detection system is required to give a response in milliseconds. This criterion needs to be taken into account in the modeling process in order for the system to be successfully implemented. In order to address the aforementioned challenges in credit card fraud detection, the plan is to account for its practical aspects in extending state-of-the-art machine learning techniques. This includes defining appropriate features and a cost-sensitive measure to be used for evaluation and training. Solutions found for specific problems in credit card fraud detection may be, at a later stage, generalized to other areas of machine learning.
1.3 SIGNIFICANCE OF THE PROJECT
This has developed the need to create a system that detects credit card fraud using FDAs. The system will facilitate an analysis of synthetic credit card transactions to determine if the transaction is genuine or fraudulent. Once determined, the system will produce an output indicating why the transaction is genuine or fraudulent based on the features within the data. By producing a system for detecting credit card fraud, the transaction results can be presented can be used as a novel method for detecting credit card fraud.
1.4 SCOPE OF THE PROJECT
There are many types of fraud, and the project scope lies within the subset of financial fraud. The project proposes to analyse transactions using FDAs and velocity checks, the algorithm presented in the project focuses specifically on detecting credit and debit card fraud because the focus of detecting fraud altogether is far too broad for an algorithm to learnfrom (Berk, R.J, 2010). Instead, traditional and modern credit card fraud.
|
1.5 PURPOSE OF THE PROJECT
The main purpose of this work is:
- For Fraud detection in real-time
- Deeper knowledge about customer behavior
- Improvement of payment data credibility
1.6 AIM AND OBJECTIVES
Aim
This project aims to develop a system that detects credit card fraud using FDAs.
Objective
- Conduct further research of traditional and modern carding techniques,
- Present current statistics of credit card fraud,
- Create a rich picture of a modern carding technique,
- Explore the detection methods of the carding technique,
- Explain how the algorithms can be conducted to detect credit card fraud,
- Evaluate the use of the algorithms and the application of them in
Reviews
There are no reviews yet.