Abstract
Operating or riding in a vehicle is one of the most dangerous things the typical person does on a regular basis. This paper describes how one company is using new technologies and techniques to mine massive amounts of vehicle crash statistics. In 1998, the company invested in new data mart technology that opened the door to more sophisticated analysis of real world insurance claims data by vehicle, by driver, and by geographic area. This paper will discuss the new data mart and illustrate some data mining tools. Four examples will be used to illustrate how the data is being mined to promote safety and better match rates to risk. These include vehicle safety, dangerous intersections, child passenger safety, and teenage driver safety.
Volume
Winter
Page
31-56
Year
2002
Categories
Actuarial Applications and Methodologies
Data Management and Information
Data Administration, Warehousing and Design
Financial and Statistical Methods
Statistical Models and Methods
Data Mining
Actuarial Applications and Methodologies
Data Management and Information
Data Organization
Financial and Statistical Methods
Statistical Models and Methods
Regression
Business Areas
Automobile
Actuarial Applications and Methodologies
Ratemaking
Publications
Casualty Actuarial Society E-Forum
Documents
