A New Biased Estimation Method Of Linear Regression Coefficient——Integrated C-K Ridge

The Least Squares Estimation has many excellent properties among the coefficient estimation methods of the Linear Regression Model.Therefore,the Least Squares Estimation is the most influential and the most widely used coefficient estimation method. It’s the core of the regression parameter esti-mates.However,with the rapid development of computer technology, there are many complex problems needing more variables.It is inevitable that the variables we select are unnecessary, unsuitable and even completely wrong. As a result,the model we established will have drawbacks,such as variables serial-correlation and multi-collincarity etc.When these situations arc occur, the estimated parameter values via the Least Squares Estimation will not sta-ble. In order to avoid this problems,statisticians improve the Least Squares Estimation and even create new coefficient estimation methods.It is the origin of biased estimates.The purpose of this paper is to solve multi-collincarity of variables.I propose a new biased estimation to estimate the parameter of the Linear Re-grcssion Model. This method is based on Ridge Estimation and Stein Estima-tion.Firstly, I will define the new estimation method.Secondly, I will proof the properties of the new method.The properties include the Linearity,the Bias,the Compression,the Optimal Property and the Admissibility. Finally,I will calculate the value of the parameter c and K.In the end of this paper, I will verify the Optimal Properties of the In-tegrated c-K Ridge Estimation Method combined with a practical problem. According to the previous theory and experience, I select seven explain vari-ables and one explained variable. Moreover, I collect a lot of data from the Website of the Bureau of Statistics. Then I suppose Linear Regression Model. Using the new Integrated c-K Ridge Estimate Method,I calculate the value of the regression parameter.Besides, I proof the good properties of the new method.