The Influence Analysis Of Ridge Estimation In Restricted Linear Model

The research about least square estimation of regression parameters in the linear model has included integrated and systemic results. This article mainly studies the influence analysis of ridge estimation in restricted linear model. Based on the proposed conditional ridge estimation, we analyse the influence of the three kinds of disturbance on the conditional ridge estimation, and give the relevant conclusions. Also the definition of generalized Cook distance and Welsch-Kuh statistics are given to measure the degree of the influence, and give the corresponding formulas.This paper introduces the history of the development and research status about general linear model and influence analysis in the first part. Then the knowledge of the matrix, ridge estimation and several lemmas are introduced. In the third and fourth chapters, we study the influence analysis of the covariance matrix disturbance and data deleted disturbance on the linear model with constraints and we establish the relationships of the conditional ridge estimation between before disturbance and after disturbance, also study the mixed influence by the covariance matrix disturbance and data deleted disturbance on the conditional ridge estimator. The fifth chapter discusses the influence of mean shift disturbance to conditional ridge estimation in restricted linear model. In the end, we give the generalized Cook distance and the Welsch-Kuh statistic to measure the degree of the influence, and an example are given.