A Variable Clustering Method and its Applying to Variable Selection in Multiple Linear Regression Model

摘要：

In this paper, we discussed a variable selection method based on hierarchical variable clustering analysis, which is different from convensional methods. In this kind of clustering method, we defined the similarity measure between two sets of variables by their first canonical correlation coefficient. And we applied this method to select variables in multiple linear regression model by an example.

关键词： variable selection clustering analysis collinear data canonical correlation.

作者: XuYuan

作者单位: School of Mathematics and Statistics, Hainan Normal University, Haikou, 571158, China

会议类型: 国际会议

会议名称: 2010 International Conference on Probability and Statistics of the International Institute for General Systems Studies(国际一般系统理论研究会中国概率统计学会第二届学术会议IIGSS-CPS2010)

会议地点: 南京

会议语种:英文

页码: 552-555

在线出版日期: 2010-07-29（万方平台首次上网日期，不代表论文的发表时间）