Unsupervised models for fraud detection

prakash_sridhar

Hi,

Has anybody come across the use of cluster analysis/other unsupervised models for fraud detection. I'm facing a specific challenge to use an unsupervised model for detecting fraud. The dataset does not have a specific variable as a target/dependent variable. I've never worked with this kind of a setup before.

This is in the field of Health Insurance. Types of fraud for example, would be one doctor billing different from the others and commiting fraud etc. I'm unable to divulge too many details in this regard. But the kind of variables we have here are similar to what we encounter in Banking - Credit Card Fraud etc...

Thanks in advance for your help

Prakash

sbutler

You could build baseline model/s then detect anomalies that differ from the baseline, as in:

Chris Curry, Robert L. Grossman, David Locke, Steve Vejcik, Joseph Bugajski: Detecting changes in large data sets of payment card data: a case study. KDD 2007: 1018-1022

--
Shane Butler
http://www.DataMiningDownUnder.com

prakash_sridhar

Thank You Shane. That was really helpful. The concept was very interesting and something we can try out in Insurance.

However, it would help if we could get more articles related to these kind of studies. 

Thanks