A Class Of Biased Estimation Improvement Research

Linear regression model is a statistical analysis of the important one of the models, and themodel parameter estimation theory and examples of analysis and linear model hot topic, thisarticle mainly aims at the problems of this kind in the partial unfolding studied,The main work ofthis paper are as follows: through to the k-D estimation improvement research, constructs abroader class of estimators, and in the mean square error and some estimates were compared inthis paper. Through argumentation, estimation is superior to other estimates of the condition, andthe condition for application to specific estimation, draw on several common estimation animportant corollary.through the improvement of generalized ridge estimation research, discussesanother to the least squares estimation improvement method, found a constrained generalizedridge estimation method of measuring, and practical examples to explain this method producesand the application process, and also to verify the rationality of this method.The estimationtheory to data analysis and examples show very little of this situation, this article in has a largecollection of data on the basis of existing estimation theory to undertake systems analysis andlots of data simulation, validation of ridge estimate, generalized ridge estimation in themean-square error matrix is better than the least squares estimate condition the feasibility andLiu estimation and the conditional root squares estimation in the MSE is superior to least squaresestimate conditions practical. Through the analysis of the data obtained for the effectiveness ofestimation method of. Also validate the rationality of this method.