基于Logistic变换的回归正交试验模型  

Regression orthogonal test model with Logistic transform

作　　者：任明杰[1] 李海燕[1] 刘洪[1] 

机构地区：[1]辽宁科技大学理学院,辽宁鞍山114051

出　　处：《辽宁科技大学学报》2012年第3期239-243,共5页Journal of University of Science and Technology Liaoning

基　　金：辽宁省自然科学基金项目(20102097)

摘　　要：在解决线性回归问题时,回归正交试验具有较大的优越性,却无法解决非线性回归问题。本文依据广义线性模型的基本原理,针对非线性回归问题,引入Logistic变换,给出了基于Logistic变换的回归正交试验模型,实现了回归正交试验在非线性回归问题中的运用。将此方法运用到车险续保问题中,实例分析说明了该模型的可行性和有效性。The regression orthogonal test has more advantages in the solution of the linear regression problems, however it can not solve the problem of nonlinear regression. In this paper, according to the basic principle of generalized linear model, in the light of nonlinear regression problem, the re- gression orthogonal test model is given besed on Logistic transform, the regression orthogonal test is realized in the application of nonlinear regression problem, and it is applied to the car insurance renew- al problem. An example analysis shows that the model is feasible and effective.

关 键 词：Logistic变换 回归正交试验 车险续保率 

分 类 号：O212.6[理学—概率论与数理统计]