2、及其应用。建模过程中的数学运算采用数学软件SPSS和Matlab进行运算。关键词：多元线性回归；模型检验；后退法；预测MultiplelinearregressionmodelanditsapplicationMengxiangmei(CollegeofmathematicalandStatisticalSciences，Statistics，Class1002，20102111977)Abstract:this article analyzes the multivariate linear regression model and its applicati
3、on, focusing on the multiple linear regression model prediction. First introduces the model of multivariate linear regression model of step roughly model, basic assumptions, inspection, and prediction ability of the model. , in the process of the establishment of the model test is th
4、e core of the modeling, model testing including fitting test, F test and t test. If has failed t test on initial model, based on the method of eliminating backward without significant variables, how far to establish linear regression model. Then, based on the 2005, 2005 China's 31 pr
5、ovinces, municipalities and autonomous regions of fiscal expenditure data and data across our country gross domestic product in 2005, set up multiple linear regression model, predict our country gross domestic product in 2006, compared with the actual data and forecast data analysis.
6、 Try to understand the multiple linear regression models and its application by example analysis .Mathematical modeling in the process of using mathematics software SPSS and MATLAB operations.Keywords:Multiplelinearregression；Modelchecking；Backwardelimination;Forecast1引言回归是社会科学定量研究方法
