NAG: Neural Feature Aggregation Framework for Credit Card Fraud Detection

Dastidar, Kanishka Ghosh; Jurgovsky, Johannes; Siblini, Wissam; He-Guelton, Liyun

doi:10.1109/icdm50108.2020.00018

Cited by 7 publications

(11 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For this, we use Word2Vec applied to the sequences of discrete values taken by the feature for all card-holders as proposed in Russac et al (2018). Each value of a categorical feature f is finally translated into a vector Ghosh et al, 2020).…”

Section: Preprocessing and Feature Engineeringmentioning

confidence: 99%

“…Besides the raw features, we additionally use aggregated features (Bahnsen et al, 2016;Jurgovsky et al, 2018;Ghosh et al, 2020). This serves as a feature engineering method to capture the past context of a card-holder.…”

Section: Preprocessing and Feature Engineeringmentioning

confidence: 99%

“…In practice, discriminating accurately between the two classes based solely on raw information about the transaction is extremely difficult. What strongly characterizes a transaction as fraudulent is its discrepancy with respect to the consumer's habits, hence the need for contextualization with aggregation or representation of the cardholder's history of purchases (Bahnsen et al, 2016;Ghosh et al, 2020;Jurgovsky et al, 2018). To the best of our knowledge, available research studies on credit card fraud detection implement similar experimental protocols with a temporal split between training data and validation data, in which each sample is a transaction, sometimes contextualized with the past transactions of the cardholder.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

The Importance of Future Information in Credit Card Fraud Detection

Nguyen¹,

Dastidar²,

Siblini³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Fraud detection systems (FDS) mainly perform two tasks: (i) real-time detection while the payment is being processed and (ii) posterior detection to block the card retrospectively and avoid further frauds. Since human verification is often necessary and the payment processing time is limited, the second task manages the largest volume of transactions. In the literature, fraud detection challenges and algorithms performance are widely studied but the very formulation of the problem is never disrupted: it aims at predicting if a transaction is fraudulent based on its characteristics and the past transactions of the cardholder. Yet, in posterior detection, verification often takes days, so new payments on the card become available before a decision is taken. This is our motivation to propose a new paradigm: posterior fraud detection with "future" information. We start by providing evidence of the ontime availability of subsequent transactions, usable as extra context to improve detection. We then design a Bidirectional LSTM to make use of these transactions. On a realworld dataset with over 30 million transactions, it achieves higher performance than a regular LSTM, which is the state-of-the-art classifier for fraud detection that only uses the past context. We also introduce new metrics to show that the proposal catches more frauds, more compromised cards, and based on their earliest frauds. We believe that fu-

show abstract

Section: Preprocessing and Feature Engineeringmentioning

confidence: 99%

Section: Preprocessing and Feature Engineeringmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The Importance of Future Information in Credit Card Fraud Detection

Nguyen¹,

Dastidar²,

Siblini³

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…These observations are so unique as to arouse the suspicion that they were generated owing to illegal acts or undetected errors. To reveal the critical information in them, many outlier detection technologies have been studied and used in various applications, such as fraud detection in credit card transactions [8,11,39] or taxes [17,19], identifying false information on e-commerce platforms [12,30] or social media [23,40,42], intrusion detection in network service requests [35,38], and detecting abnormal trajectories during traic monitoring [4,34].…”

Section: Introductionmentioning

confidence: 99%

Dual-MGAN: An Efficient Approach for Semi-supervised Outlier Detection with Few Identified Anomalies

Sun

Liu

et al. 2022

ACM Trans. Knowl. Discov. Data

View full text Add to dashboard Cite

Outlier detection is an important task in data mining, and many technologies for it have been explored in various applications. However, owing to the default assumption that outliers are not concentrated, unsupervised outlier detection may not correctly identify group anomalies with higher levels of density. Although high detection rates and optimal parameters can usually be achieved by using supervised outlier detection, obtaining a sufficient number of correct labels is a time-consuming task. To solve these problems, we focus on semi-supervised outlier detection with few identified anomalies and a large amount of unlabeled data. The task of semi-supervised outlier detection is first decomposed into the detection of discrete anomalies and that of partially identified group anomalies, and a distribution construction sub-module and a data augmentation sub-module are then proposed to identify them, respectively. In this way, the dual multiple generative adversarial networks (Dual-MGAN) that combines the two sub-modules can identify discrete as well as partially identified group anomalies. In addition, in view of the difficulty of determining the stop node of training, two evaluation indicators are introduced to evaluate the training status of the sub-GANs. Extensive experiments on synthetic and real-world data show that the proposed Dual-MGAN can significantly improve the accuracy of outlier detection, and the proposed evaluation indicators can reflect the training status of the sub-GANs.

show abstract

“…The traditional paradigm of fraud detection solutions in financial institutions consists of formulating it as a classification problem with a focus on improving the fraud recall rates of these classification models. Several papers in the literature have proposed different methods for creating robust features that are immune to model/concept drift using statistical, deep learning, and unsupervised techniques ( [27], [4], [46], [2], [9]). However, these methods of problem formulation ignore a number of issues:…”

Section: Introductionmentioning

confidence: 99%

Application of Deep Reinforcement Learning to Payment Fraud

Siddharth¹,

Kayathwal²,

Wadhwa³

et al. 2021

Preprint

View full text Add to dashboard Cite

The large variety of digital payment choices available to consumers today has been a key driver of e-commerce transactions in the past decade. Unfortunately, this has also given rise to cybercriminals and fraudsters who are constantly looking for vulnerabilities in these systems by deploying increasingly sophisticated fraud attacks. A typical fraud detection system employs standard supervised learning methods where the focus is on maximizing the fraud recall rate. However, we argue that such a formulation can lead to sub-optimal solutions. The design requirements for these fraud models requires that they are robust to the high-class imbalance in the data, adaptive to changes in fraud patterns, maintain a balance between the fraud rate and the decline rate to maximize revenue, and be amenable to asynchronous feedback since usually there is a significant lag between the transaction and the fraud realization. To achieve this, we formulate fraud detection as a sequential decision-making problem by including the utility maximization within the model in the form of the reward function. The historical decline rate and fraud rate define the state of the system with a binary action space composed of approving or declining the transaction. In this study, we primarily focus on utility maximization and explore different reward functions to this end. The performance of the proposed Reinforcement Learning system has been evaluated for two publicly available fraud datasets using Deep Q-learning and compared with different classifiers. We aim to address the rest of the issues in future work. CCS CONCEPTS• Applied computing → Online banking; Secure online transactions; Online shopping.

show abstract

NAG: Neural Feature Aggregation Framework for Credit Card Fraud Detection

Cited by 7 publications

References 14 publications

The Importance of Future Information in Credit Card Fraud Detection

The Importance of Future Information in Credit Card Fraud Detection

Dual-MGAN: An Efficient Approach for Semi-supervised Outlier Detection with Few Identified Anomalies

Application of Deep Reinforcement Learning to Payment Fraud

Contact Info

Product

Resources

About