Machine Learning Advances in Payment Card Fraud Detection provides a thorough review of the state-of-the-art in fraud detection research that is ideal for graduate-level readers and professionals. Through a comprehensive examination of fraud analytics that covers data collection, steps for cleaning and processing data, tools for analysing data, and ways to draw insights, the book argues for a new direction to be taken in developing state-of the-art payment fraud detection techniques. It uses an extensive analysis and description of an exemplar fraud detection algorithm, SOAR, to illustrate how a detailed understanding of the payment fraud domain can be used to motivate further advances in fraud detection techniques. The book concludes with a discussion of opportunities for future research, such as developing holistic approaches for countering fraud.
- Provides a detailed analysis of the payment fraud detection domain
- Develops an evaluation methodology for fraud detection techniques that takes full account of the business needs of the payment card industry
- Introduces state-of the-art payment fraud detection techniques
Table of Contents
1. Introduction2. History of payment cards
3. Growth of payment card fraud
4. The fraud detection problem
5. Understanding the fraud domain
6. Cost of payment fraud
7. History of payment card fraud detection methods
8. The pivotal event and disruptive technologies
9. The Sparse Oracle-based Adaptive Rule extraction algorithm
10. Real-world data empirical evaluation
11. Discussion and conclusion
