A Linear Regression Model Of Biased Estimation To Improve Research

The linear regression model theory is a very important statistical theory in modern times, which and it has been widely applied in all fields. It plays a very important role in the processing of data, and it has been widely used in many fields such as natural science and social economy, etc. In fact, it expresses a variety of factors that have some correlations in actual problems with mathematical relationship. The basic theory and the basic methods of the linear regression model provides basic tools for other statistical methods and statistical theory. Due to the comprehensive application of the linear regression model in various aspects, experts and scholars have done much research and improvement to make it more accurate and simple.This paper first introduces the present development situation and the several common basic knowledge of biased estimate for prove that a certain foundation for later.After that, this paper studys the determinant parameters in the biased estimate. According to the LIU-type Principal component estimate its parameter are studied, and a kind of biased estimate is obtained. Then this paper explored some properties of it and further explored the superiority that this kind of estimate may have in a certain conditions.Finally, according to the conclusion obtained above, we further studied and discussed some important properties of the new estimate, making use of the mean-square error and the mean-square error matrix means to study what condition the new estimation superior to least squares estimate and principal components estimate.we explore the admissible property that the new estimate has.