Search results

Purpose: This chapter aims to examine machine learning (ML) models for predicting credit card fraud (CCF).

Need for the study: With the advance of technology, the world is increasingly relying on credit cards rather than cash in daily life. This creates a slew of new opportunities for fraudulent individuals to abuse these cards. As of December 2020, global card losses reached $28.65billion, up 2.9% from $27.85 billion in 2018, according to the Nilson 2019 research. To safeguard the safety of credit card users, the credit card issuer should include a service that protects customers from potential risks. CCF has become a severe threat as internet buying has grown. To this goal, various studies in the field of automatic and real-time fraud detection are required. Due to their advantageous properties, the most recent ones employ a variety of ML algorithms and techniques to construct a well-fitting model to detect fraudulent transactions. When it comes to recognising credit card risk is huge and high-dimensional data, feature selection (FS) is critical for improving classification accuracy and fraud detection.

Methodology/design/approach: The objectives of this chapter are to construct a new model for credit card fraud detection (CCFD) based on principal component analysis (PCA) for FS and using supervised ML techniques such as K-nearest neighbour (KNN), ridge classifier, gradient boosting, quadratic discriminant analysis, AdaBoost, and random forest for classification of fraudulent and legitimate transactions. When compared to earlier experiments, the suggested approach demonstrates a high capacity for detecting fraudulent transactions. To be more precise, our model’s resilience is constructed by integrating the power of PCA for determining the most useful predictive features. The experimental analysis was performed on German credit card and Taiwan credit card data sets.

Findings: The experimental findings revealed that the KNN achieved an accuracy of 96.29%, recall of 100%, and precision of 96.29%, which is the best performing model on the German data set. While the ridge classifier was the best performing model on Taiwan Credit data with an accuracy of 81.75%, recall of 34.89, and precision of 66.61%.

Practical implications: The poor performance of the models on the Taiwan data revealed that it is an imbalanced credit card data set. The comparison of our proposed models with state-of-the-art credit card ML models showed that our results were competitive.

The best algorithm that was implemented on this Brazilian dataset was artificial immune system (AIS) algorithm. But the time and cost of this algorithm are high. Using asexual reproduction optimization (ARO) algorithm, the authors achieved better results in less time. So the authors achieved less cost in a shorter time. Their framework addressed the problems such as high costs and training time in credit card fraud detection. This simple and effective approach has achieved better results than the best techniques implemented on our dataset so far. The purpose of this paper is to detect credit card fraud using ARO.

In this paper, the authors used ARO algorithm to classify the bank transactions into fraud and legitimate. ARO is taken from asexual reproduction. Asexual reproduction refers to a kind of production in which one parent produces offspring identical to herself. In ARO algorithm, an individual is shown by a vector of variables. Each variable is considered as a chromosome. A binary string represents a chromosome consisted of genes. It is supposed that every generated answer exists in the environment, and because of limited resources, only the best solution can remain alive. The algorithm starts with a random individual in the answer scope. This parent reproduces the offspring named bud. Either the parent or the offspring can survive. In this competition, the one which outperforms in fitness function remains alive. If the offspring has suitable performance, it will be the next parent, and the current parent becomes obsolete. Otherwise, the offspring perishes, and the present parent survives. The algorithm recurs until the stop condition occurs.

Results showed that ARO had increased the AUC (i.e. area under a receiver operating characteristic (ROC) curve), sensitivity, precision, specificity and accuracy by 13%, 25%, 56%, 3% and 3%, in comparison with AIS, respectively. The authors achieved a high precision value indicating that if ARO detects a record as a fraud, with a high probability, it is a fraud one. Supporting a real-time fraud detection system is another vital issue. ARO outperforms AIS not only in the mentioned criteria, but also decreases the training time by 75% in comparison with the AIS, which is a significant figure.

In this paper, the authors implemented the ARO in credit card fraud detection. The authors compared the results with those of the AIS, which was one of the best methods ever implemented on the benchmark dataset. The chief focus of the fraud detection studies is finding the algorithms that can detect legal transactions from the fraudulent ones with high detection accuracy in the shortest time and at a low cost. That ARO meets all these demands.

This paper aims to provide information of parameters and techniques used in the automated fraud detection system during online transaction. With the increase in the use of online transactions, the concerns regarding data security have also increased. To tackle the frauds, lot of research has been done and plethora of papers are available on the related topics. The purpose of this paper is to provide the clear pathway for researchers to move in the direction of development of automated fraud detection system to prevent the fraud during online transaction.

This literature review analyses and compares the different types of techniques for detecting fraud during online transaction. An in-depth study of the most prominent journals has been done and the core methodology of the papers has been presented. This article also shed some light on different types of parameters used in fraud detection techniques during online transaction.

There are vast varieties of various fraud detection techniques, and every technique has completed task in its own way. After studying approximately 41 research papers, 14 books and four reports, in total 30 parameters have been identified and a detailed study of the parameters has been presented. The parameters are also listed with their details that how these parameters are used in the security system for detecting online transaction fraud.

This paper provides empirical insight about the parameters and their prominence in the development of automated fraud detection security system of online transaction. This paper encourages the researchers to development of improved fraud detection system.

This paper will pave the way for researchers to do a focused research on the fraud detection methodologies. The analysis will help in zeroing down the most prevalent topic of research in this field. The researchers will be able to understand the internal details of parameters and techniques used in the fraud detection systems. This literature also helps the research to think in a variety of ways that how these parameters will be used in the development of fraud detection system.

This paper is one of the most comprehensive reviews in its field. It tries and attempts to fill a void created because of lack of compilation of the laid fraud detection parameters.

Fraud transaction detection has become a significant factor in the communication technologies and electronic commerce systems, as it affects the usage of electronic payment. Even though, various fraud detection methods are developed, enhancing the performance of electronic payment by detecting the fraudsters results in a great challenge in the bank transaction.

This paper aims to design the fraud detection mechanism using the proposed Harris water optimization-based deep recurrent neural network (HWO-based deep RNN). The proposed fraud detection strategy includes three different phases, namely, pre-processing, feature selection and fraud detection. Initially, the input transactional data is subjected to the pre-processing phase, where the data is pre-processed using the Box-Cox transformation to remove the redundant and noise values from data. The pre-processed data is passed to the feature selection phase, where the essential and the suitable features are selected using the wrapper model. The selected feature makes the classifier to perform better detection performance. Finally, the selected features are fed to the detection phase, where the deep recurrent neural network classifier is used to achieve the fraud detection process such that the training process of the classifier is done by the proposed Harris water optimization algorithm, which is the integration of water wave optimization and Harris hawks optimization.

Moreover, the proposed HWO-based deep RNN obtained better performance in terms of the metrics, such as accuracy, sensitivity and specificity with the values of 0.9192, 0.7642 and 0.9943.

An effective fraud detection method named HWO-based deep RNN is designed to detect the frauds in the bank transaction. The optimal features selected using the wrapper model enable the classifier to find fraudulent activities more efficiently. However, the accurate detection result is evaluated through the optimization model based on the fitness measure such that the function with the minimal error value is declared as the best solution, as it yields better detection results.

This chapter introduces a risk control framework on credit card fraud instead of providing a solely binary classifier model. The anomaly detection approach is adopted to identify fraud events as the outliers of the reconstruction error of a trained autoencoder (AE). The trained AE shows fitness and robustness on the normal transactions and heterogeneous behavior on fraud activities. The cost of false-positive normal transactions is controlled, and the loss of false-negative frauds can be evaluated by the thresholds from the percentiles of reconstruction error of trained AE on normal transactions. To align the risk assessment of the economic and financial situation, the risk manager can adjust the threshold to meet the risk control requirements. Using the 95th percentile as the threshold, the rate of wrongly detecting normal transactions is controlled at 5% and the true positive rate is 86%. For the 99th percentile threshold, the well-controlled false positive rate is around 1% and 83% for the truly detecting fraud activities. The performance of a false positive rate and the true positive rate is competitive with other supervised learning algorithms.

Currency usage either in the physical or electronic marketplace through chip-based or magnetic strip-based plastic card becoming the vulnerable point for the handlers. Proper education and awareness can only thrive when concrete fraud detection techniques are being suggested together with potential mitigation possibilities. The purpose of this research study is tendering in the same direction with a suitable plan of action in developing the authentication strength metric to give weightage marks for authentication techniques.

In this research study, a qualitative in-depth exploration approach is being adapted for a better description, interpretation, conceptualization for attaining exhaustive insights into specific notions. A concrete method of observation is being adopted to study various time boxed reports on plastic card fraud and its possible impacts. Content and narrative analysis are being followed to interpret more qualitative and less quantitative story about existing fraud detection techniques. Moreover, an authentication strength metric is being developed on the basis of time, cost and human interactions.

The archived data narrated in various published research articles represent the local and global environment and the need for plastic card money. It gives the breathing sense and capabilities in the marketplace. The authentication strength metric gives a supporting hand for more solidification of the authentication technique with respect to the time, cost and human ease.

The research study is well controlled and sufficient interpretive. The empirical representation of authentication technique and fraud detection technique identification and suggestive mitigation gives this research study an implication view for the imbibing research youths. An application and metric based pathway of this research study provides a smoother way to tackle futuristic issues and challenges.

This research study represents comprehensive knowledge about the causes of the notion of plastic card fraud. The authentication strength metric represents the novelty of a research study which produced on the basis of rigorous documentary and classified research analysis. The creativity of the research study is rendering the profound and thoughtful reflection of the novel dimension in the same domain.

Purpose: The study aims to explore the wider acceptance of blockchain technology and growing faith in this technology among all business domains to mitigate the chances of fraud in various sectors.

Design/Methodology/Approach: The authors focus on studies conducted during 2015–2022 using keywords such as blockchain, fraud detection and financial domain for Systematic Literature Review (SLR). The SLR approach entails two databases, namely, Scopus and IEEE Xplore, to seek relevant articles covering the effectiveness of blockchain technology in controlling financial fraud.

Findings: The findings of the research explored different types of business domains using blockchains in detecting fraud. They examined their effectiveness in other sectors such as insurance, banks, online transactions, real estate, credit card usage, etc.

Practical Implications: The results of this research highlight (1) the real-life applications of blockchain technology to secure the gateway for online transactions; (2) people from diverse backgrounds with different business objectives can strongly rely on blockchains to prevent fraud.

Originality/Value: The SLR conducted in this study assists in the identification of future avenues with practical implications, making researchers aware of the work so far carried out for checking the effectiveness of blockchain; however, it does not ignore the possibility of zero to less effectiveness in some businesses which is yet to be explored.

The purpose of this paper is to discuss credit card fraud in Trinidad and Tobago.

The paper describes credit card typologies in Trinidad and Tobago and the existing law governing such fraud. It outlines the success and inadequacies of the enforcement machinery and issues involving credit card fraud detection and prevention.

The law regarding credit cards is in a very confused and unsatisfactory state in Trinidad and Tobago. Education in counteracting the criminal activities of credit card fraudsters is vital. Informing the public of the various fraudulent typologies relative to credit cards and at the same time, advising members how to protect themselves are the most effective methods to address the fraud problem. The Bankers Association of Trinidad and Tobago should also play a critical role in addressing credit card fraud. The association should formulate credit card policies along similar principles as those formulated for cheque fraud to benefit all banks and merchants.

Credit card fraud has only been a recent phenomenon in Trinidad and Tobago. This paper is valuable in offering suggestions as to the way forward in the prevention of such fraud.

This paper aims to reviews the literature on applying visualization techniques to detect credit card fraud (CCF) and suspicious money laundering transactions.

In surveying the literature on visual fraud detection in these two domains, this paper reviews: the current use of visualization techniques, the variations of visual analytics used and the challenges of these techniques.

The findings reveal how visual analytics is used to detect outliers in CCF detection and identify links to criminal networks in money laundering transactions. Graph methodology and unsupervised clustering analyses are the most dominant types of visual analytics used for CCF detection. In contrast, network and graph analytics are heavily used in identifying criminal relationships in money laundering transactions.

Some common challenges in using visualization techniques to identify fraudulent transactions in both domains relate to data complexity and fraudsters’ ability to evade monitoring mechanisms.

This paper aims to present a bibliometric analysis of publications from the Scopus database on fraud detection studies.

The current research used Microsoft Excel to conduct the frequency analysis, VOSviewer for data visualisation and Harzing’s Publish or Perish for citation metrics and analysis.

In alignment with these research results, the publications on fraud detection studies have been consistently increasing since 2005. India was rated first as the most active country in fraud detection research. Tongji University from China was the most active institution that published significant publications related to fraud detection research. A total of 160 scholars from 89 various countries and 160 different institutions published several fraud detections studies with multi-authors’ participation in different languages.

To the best of the authors knowledge, this study is the first study to review fraud detections publications in the Scopus science database.