Paper Title
Regression Analysis of Car Insurance Selection Model an Appropriate or Non-Appropriate Method

Abstract
In insurance sector sales of the particular policies are always depends on many parameters. Significant past data are available in term of quantitative and qualitative form of sales. The present paper discuss the car policy sales predication model with ten input parameters namely, age of policy holder, job of policy holder, marital status of policy holder, education of policy holder, balance in policy holder account, housing loan, car loan, last contact month, number times customer contacted by insurance agent, and time of contact duration, etc. The information of 1000 customers used to develop this model. The data is normalized and appropriate weights are given to the input parameters. The multivariable regression model used to evaluate the constants. The best suitable car insurance model is predicated and the critical parameters for the car insurance sales are identified. The analysis suggest that customer job, education of the customer and existing car loan are the critical parameters for the car sales. The car insurance models developed to predict the accurate car insurance sales with maximum error are8. 88 %. The change in the critical limit factor from 0.2 to 0.23 is used to predict number of car policies to be sold and also help to insurance provider to identify the probable customer. Keywords - Car Insurance, Predication Model, Regression Analysis, Insurance