PLS-Logistic Regression on Functional Data

摘要：

Regression modeling in many fields, such as credit rating, banking industry and macroeconomic studies, is an important approach. However, MulticoUinearity in the independent variable sets is harmful to Ordinary Least Squares (OLS) Regression. Partial Least Squares (PLS) Regression enables modeling under the condition of multicollinearity. In the fields of Credit Rating, many independent variables are related functional data, and the dependent variable is a categorical variable. For these problems, Functional PLS-Logistic Regression provides an approach of building regression model under the condition of multicollinearity. Empirical study shows that the GDP per capita of provinces in China has an obvious distribution feature which ensures the reasonability of classify the provinces according to their geography locations.

关键词： PLS Regression Multicollinearity Functional data Logistic Regression

作者: Jie Wang Shengshuai Wang Kefei Huang Ying Li

作者单位: Dagong Global Credit Rating Co., Ltd., Beijing, China Beijing University of Aeronautics and Astronautics, Beijing, China Tiantan Hospital, Beijing, China Canada International School, Beijing, China

会议类型: 国际会议

会议名称: The 6th International Conference on Partial Least Squares and Related Methods(第六届偏最小二乘及相关方法国际会议)

会议地点: 北京

会议语种:英文

页码: 71-76

在线出版日期: 2009-09-04（万方平台首次上网日期，不代表论文的发表时间）