International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No.

Supervised Machine Learning Model for Credit Card Fraud Detection

Manoj Aryal, Prof. Anurag jain and Dr. A.K. Sachan Deptt. Of Computer science & Engineering, Radharaman Institute of Technology & science, Bhopal-462051, India manoj_aryal_32@yahoo.co.in Abstract
Electronic transactions with credit card become more popular mode of payment nowadays such as online shopping and banking. Credit card frauds increased with the introduction of latest transaction technologies. To detect credit card frauds in electronic transactions becomes the focus of risk control of banks. Several researchers have proposed their work for credit card fraud detection. We proposed a binary classification model for credit card fraud detection. This Binary Classification model gives the accuracy up to 97% and also reduces the false alarm rate. . Keywords: Fraud detection, RBF, Credit card, True Positive

1. Introduction
Credit card fraud is a wide-ranging term for theft and fraud committed using a credit card or any similar payment mechanism as a fraudulent source of funds in a transaction. The purpose may be to obtain goods without paying, or to obtain unauthorized funds from an account [1]. Credit card fraud detection has drawn a lot of research interest and a number of techniques, with special emphasis on data mining and neural networks, have been suggested. Ghosh and Reilly [2] have proposed credit card fraud detection with a neural network. They have built a detection system, which is trained on a large sample of labeled credit card account transactions. These transactions contain example fraud cases due to lost cards, stolen cards, application fraud, counterfeit fraud, mail-order fraud, and non received issue (NRI) fraud.

1.1 Types of Credit Card Fraud [3]

Credit fraud can fall into one of five categories: a. b. c. d. Counterfeit credit card Lost or Stolen Cards No-Card Fraud Non-Receipt Fraud

55

International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No. 8

e.

Identity Theft Fraud

a.

Counterfeit credit card:-Makes up for 37% of all funds lost through credit card frauds. To make fake cards criminals use the newest technology to skim information contained on magnetic stripes of cards and to pass security features such as holograms.

b.

Lost or Stolen Cards:- Cards stolen from their cardholders or lost by them account for 23% of all card frauds. Often, cards are stolen from the workplace, gym, and unattended vehicles.

c.

No-Card Fraud:- Comprise 10% of all the losses and is completed without the physical card in hand. This can happen by giving your credit card information on the phone to shady telemarketers and deceptive Internet sites that are promoting the sales of their non-existent goods and services.

d.

Non-Receipt Fraud:- Non-Receipt Fraud is responsible for 7% of all losses. It occurs when new or replaced cards mailed by your card company are stolen during the process of being mailed. However, this type of fraud is on the decline with the card-activation process that most companies use.

e.

Identity Theft Fraud:- Accounts for 4% of all losses, and occurs when criminals apply for a card using someone else identity and information.

2. RELATED WORK
All Recently, Syeda et al [4] have used parallel granular neural networks (PGNNs) for improving the speed of data mining and knowledge discovery process in credit card fraud detection. A complete system has been implemented for this purpose. Stolfo et al. [5] suggest a credit card fraud detection system (FDS) using Meta learning techniques to learn models of fraudulent credit card transactions. Meta learning is a general strategy that provides a means for combining and integrating a number of separately built classifiers or models. A Meta classifier is thus trained on the correlation of the predictions of the base classifiers. Aleskerov et al. [6] present CARDWATCH, a database mining system used for credit card fraud detection. The system, based on a neural learning module, provides an interface to a variety of commercial databases. Kim and Kim have identified 56

International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No. 8

skewed distribution of data and mix of legitimate and fraudulent transactions as the two main reasons for the complexity of credit card fraud detection. [7] Based on this observation, they use fraud density of real transaction data as a confidence value and generate the weighted fraud score to reduce the number of misdetections. Brause et al [8] have developed an approach that involves advanced data mining techniques and neural network algorithms to obtain high fraud coverage. Chiu and Tsai [9] have proposed Web services and data mining techniques to establish a collaborative scheme for fraud detection in the banking industry. With this scheme, participating banks share knowledge about the fraud patterns in a heterogeneous and distributed environment. To establish a smooth channel of data exchange, Web services techniques such as XML, SOAP, and WSDL are used. Phua et al. [10] have done an extensive survey of existing data-mining-based FDSs and published a comprehensive report. Prodromidis and Stolfo[11] use an agent-based approach with distributed learning for detecting frauds in credit card transactions. It is based on artificial intelligence and combines inductive learning algorithms and Meta learning methods for achieving higher accuracy. Phua et al.suggest[12] the use of meta classifier similar to in fraud detection problems. They consider native Bayesian, C4.5, and Back Propagation neural networks as the base classifiers. A Meta classifier is used to determine which classifier should be considered based on skewness of data. The problem with most of the above mentioned approaches is that they require labeled data for both genuine, as well as fraudulent transactions, to train the classifiers. Getting realworld fraud data is one of the biggest problems associated with credit card fraud detection. Also, these approaches cannot detect new kinds of frauds for which labeled data is not available. In contrast, we present a supervised machine learning technique for credit card FDS.

3. SVM (Support Vector Machine)

SVMs were developed by Cortes & Vapnik [13] for binary classification. Basically, we are looking for the optimal separating hyper plane between the two classes by maximizing the margin between the classes' closest points (see figure below) the points lying on the boundaries are called support vectors, and the middle of the margin is our optimal separating hyper plane; Data points on the wrong side of the discriminant margin are weighted down to reduce their influence (soft margin). When we cannot find a linear separator, data points are projected into an (usually) higher-dimensional space where the data points effectively become linearly separable 57

International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No. 8

(this projection is realized via kernel techniques);

Figure 1 Hyper plane with support vectors

Figure 2 Mapping from nonlinear to linear SVM

58

International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No. 8

4. Proposed Model , EXPERIMENTS & RESULTS

Here we detail the proposed model for detection of Fraud Transactions.

Figure 3 Credit Card Fraud Detection model

Because of privacy maintained by banks no real data is available for fraud detection. Hence for credit card fraud detection we have generated the transaction dataset using uniform probability distribution. We used MATLAB [14] for the implementation of algorithm because of its rich sets of mathematical function and also supporting the inbuilt functions for SVM. We trained SVM using generated transaction data set by the different Fraud probabilities. TPR = True Positive Rate TNR = True Negative Rate FPR = False Positive Rate FNR = False Negative Rate A comparision between Linear, Quadratic, RBF kernel has been performed which shows that the RBF 59

International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No. 8

kernel outperforms to Linear & quadratic kernel. It has maximum accuracy up to 97%, maximum TPR (99%), maximum TNR (98%) & maximum FPR (7%), maximum FNR (6%), it also behaves almost same for all type of data set generated (having very low fraud data & high fraud data).

5. CONCLUSION
In this paper we have proposed an application of SVM in credit card fraud detection. The different steps in credit card transaction processing are represented in figure 3. The results show that among the three different kernels in SVM. The RBF kernel performs better then the others. Supervised machine learning model gives the better results in comparison to the hidden markov model and gives accuracy up to 97%, also reducing the false alarm rate.

References [1] http://en.wikipedia.org/wiki/Credit_card_fraud

[2] S. Ghosh and D.L. Reilly, Credit Card Fraud Detection with a Neural-Network, Proc. 27th Hawaii Intl Conf. System Sciences: Information Systems: Decision Support and Knowledge-Based Systems, vol. 3, pp. 621-630, 1994. [3] [4] http://www.spamlaws.com/credit-fraud-stats.html M. Syeda, Y.Q. Zhang, and Y. Pan, Parallel Granular Networks for Fast Credit Card Fraud Detection, Proc. IEEE Intl Conf. Fuzzy Systems, pp. 572-577, 2002. [5] S.J. Stolfo, D.W. Fan, W. Lee, A.L. Prodromidis, and P.K. Chan,Credit Card Fraud Detection Using Meta-Learning: Issues and Initial Results, Proc. AAAI Workshop AI Methods in Fraud and Risk Management, pp. 83-90, 1997.

60

International Journal of Computational Intelligence and Information Security, August 2011 Vol. 2, No. 8
[6] E. Aleskerov, B. Freisleben, and B. Rao, CARDWATCH: ANeural Network Based Database Mining System for Credit Card Fraud Detection, Proc. IEEE/IAFE: Computational Intelligence for Financial Eng., pp. 220-226, 1997. [7] M.J. Kim and T.S. Kim, A Neural Classifier with Fraud Density Map for Effective Credit Card Fraud Detection, Proc. Intl Conf. Intelligent Data Eng. and Automated Learning, pp. 378-383, 2002. [8] R. Brause, T. Langsdorf, and M. Hepp, Neural Data Mining for Credit Card Fraud Detection, Proc. IEEE Intl Conf. Tools with Artificial Intelligence, pp. 103-106, and 1999. [9] C. Chiu and C. Tsai, A Web Services-Based Collaborative Scheme for Credit Card Fraud Detection, Proc. IEEE Intl Conf. e-Technology, e-Commerce and e-Service, pp. 177-181, 2004. [10] C. Phua, V. Lee, K. Smith,and R. Gayler,A Comprehensive Survey of Data Mining-Based Fraud Detection Research, http:// www.bsys.monash.edu.au/people/cphua/, Mar. 2007. [11] S. Stolfo and A.L. Prodromidis, Agent-Based Distributed Learning Applied to Fraud Detection, Technical Report, CUCS-014-99, Columbia Univ., 1999. [12] C. Phua, D. Alahakoon, and V. Lee, Minority Report in Fraud Detection: Classification of Skewed Data, ACM SIGKDD Explorations Newsletter, vol. 6, no. 1, pp. 50-59, 2004. [13] David Meyer, Support Vector Machines, Technische University at Wien, Austria, May 22, 2011.

[14] http://www.mathworks.com/products/matlab/

61