AnalyticBridge

A Data Science Central Community

Clinostats - Fraud detection in Insurance domain

Dear all,

I came across a very interesting concept called ‘Clinostats’ used by SIU team for investigation of suspicious claims to catch real fraudster. The dictionary meaning of the same is “blinkers”.

I am not very sure about the application of this method, but as per my understanding, SIU team select a very small sample of suspicious claims for final investigation to maintain a balance between customer satisfaction and fraud identification (Investigation of false positive case has inverse relation with customer satisfaction).I do understand that sample selection may depend on product line, severity of claim and also on nature of business.

I want to know if any of you came across of such a concept.

Cheers!!!!!

Shounak Ghosal

0 members like this

Replies to This Discussion

Permalink Reply by Tilmann Bruckhaus on May 15, 2010 at 8:41am

I published a chapter providing practical metrics and techniques for fighting insurance fraud with data mining:

Zhu, X. and Davidson, I. 2007 Knowledge Discovery and Data Mining: Challenges and Realities. IGI Publishing. Chapter VII: The Business Impact of Predictive Analytics. Tilmann Bruckhaus.

Let me know if you have any questions.

Permalink Reply by Dr Vadlamani Ravi on February 25, 2013 at 11:12pm