Abraham, Jon P.
Posterro, Barry John
Generalized linear models are starting to gain popularity among actuaries in most countries for target marketing analysis. In order to better understand how these models work, a project was commissioned regarding medical providers and their reimbursement ratios. By using one-way analysis, several factors were selected to model the response variable and the factors' significance was determined by using an algorithm in the statistical software, SAS. Several general linear models were set up and tested to fit the reimbursement ratio. By calculating and analyzing each correlation, we were able to find a model that matched the reimbursement ratio with an 83 percent correlation.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.