Decision Tree as a Predictive Modeling Tool for General Insurance Claims
Loading...
Date
March, 2016
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The study explored the relationship between the risk factors of policyholders of
an insurance company and the occurrence of claim. These relationships were
investigated by using decision trees developed by Breiman et al (1984) also
referred to as CART analysis. The R software (Rattle) version 0.99.483 was
used to analyze the data using the rpart component.The Analysis corroborated
the view held by Yeo et al. (2001) and Mayhew et al. (2003) that a major risk
factor that a ects the occurrence of claim is the age of policyholders and also
proved and ascertained Tranter and Warn (2008) views, that a group of young
drivers have higher risk of accident than the same group of older drivers. Most
importantly the analysis revealed that the prediction of whether a policyholder
would make a claim in the coming year depends in part and in whole on the claim
history of the policyholder.
Description
A thesis submitted to the Department of Mathematics,
Kwame Nkrumah University of Science and Technology in partial fulfillment of the requirement for the degree of MSc Actuarial Science,