Control of Bias in Principal Components Regression Estimator

摘要：

　　The principal components regression is one of effective methods to combat multicollinearity in the linear regression model.T here may be large bias in the principal components regression estimator and it is harmful for statistical inference.In this paper a testing to control large bias in the principal components regression estimator is proposed.The result is applied to theadvertising data from Ref 8.

关键词： Multicollinearity Principal components regression Control of bias Hypothesis test

作者: GAO Feng

作者单位: Faculty of Mathematical and Physics,Huaiyin Institute of Technology, P.R.China,223003

会议类型: 国际会议

会议名称: The 6th International Institute of Statistics & Management Engineering Symposium(IISMES2014) (第六届国际统计与管理工程学术研讨会)

会议地点: 山东 威海

会议语种:英文

页码: 883-887

在线出版日期: 2014-07-20（万方平台首次上网日期，不代表论文的发表时间）