FraudMiner: A Novel Credit Card Fraud Detection Model Based on Frequent Itemset Mining

This paper proposes an intelligent credit card fraud detection model for detecting fraud from highly imbalanced and anonymous credit card transaction datasets. The class imbalance problem is handled by finding legal as well as fraud transaction patterns for each customer by using frequent itemset mi...

Full description

Saved in:
Bibliographic Details
Published in:TheScientificWorld 2014-01, Vol.2014 (2014), p.1-10
Main Authors: Seeja, K. R., Zareapoor, Masoumeh
Format: Article
Language:English
Subjects:
Accounts Payable and Receivable
Algorithms
Behavior
Classification
Correlation coefficient
Credit card fraud
Data mining
Data Mining - statistics & numerical data
Datasets
Fraud
Fraud - prevention & control
Fraud prevention
Genetic algorithms
Humans
Identification and classification
Immune system
Methods
Neural networks
Pattern Recognition, Automated
Smartphones
Support vector machines
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper proposes an intelligent credit card fraud detection model for detecting fraud from highly imbalanced and anonymous credit card transaction datasets. The class imbalance problem is handled by finding legal as well as fraud transaction patterns for each customer by using frequent itemset mining. A matching algorithm is also proposed to find to which pattern (legal or fraud) the incoming transaction of a particular customer is closer and a decision is made accordingly. In order to handle the anonymous nature of the data, no preference is given to any of the attributes and each attribute is considered equally for finding the patterns. The performance evaluation of the proposed model is done on UCSD Data Mining Contest 2009 Dataset (anonymous and imbalanced) and it is found that the proposed model has very high fraud detection rate, balanced classification rate, Matthews correlation coefficient, and very less false alarm rate than other state-of-the-art classifiers.
ISSN:2356-6140
1537-744X
1537-744X
DOI:10.1155/2014/252797