I work in the health insurance field, but a general treatment of fraud detection methodologies would still be helpful.
So far I've discovered brief articles outlining particular techniques used in fraud detection, but not a comprehensive overview.
$\begingroup$
$\endgroup$
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
[Note: I wanted to simply "add a comment" here but the system is not allowing me]. I do not have particular experience with fraud detection specifically but my understanding is that it is closely related to anomaly and outlier detection and my response here reflects that understanding. Anyway, this paper provides a pretty comprehensive overview of various anomaly detection techniques, especially those prevalent in the machine-learning/pattern-recognition literature. Note, however, the paper is a little dated (from 2007) so you might want to find an additional reference for newer methods. Another potential reference might be this book.
-
1$\begingroup$ +1 Thank you, this is helpful. I'm guessing that part of the reason there are no textbooks specifically describing fraud detection is that analysts are simply using existing statistical/machine learning/data mining anomaly detection techniques off the shelf. Another reason - homebrewed methodologies, plus information about which features analysts typically model for fraud detection, may be proprietary knowledge. Companies may not want to tip off potential fraudsters by revealing their "secret sauce" for fraud detection. $\endgroup$– RobertFCommented Jul 18, 2017 at 15:18