多元回归用matlab怎么算

wxkdesky

多元回归用matlab怎么算呀？不会，有没有示例程序呀？包括求参数，求残差，求置信区间，最后怎么检验？

HNzhangjie

额   我也需要

hzs2012

给个实例，好点儿。。。意思是给个题目

hzs2012

不需要例子的话，直接百度、google就好了。很多很多

liwenhui

使用函数，regress()，具体调用方式见matlab的帮助。

2. regress Multiple linear regression using least squares.
   
3.     B = regress(Y,X) returns the vector B of regression coefficients in the
   
4.     linear model Y = X*B.  X is an n-by-p design matrix, with rows
   
5.     corresponding to observations and columns to predictor variables.  Y is
   
6.     an n-by-1 vector of response observations.
   
7. 
   
8.     [B,BINT] = regress(Y,X) returns a matrix BINT of 95% confidence
   
9.     intervals for B.
   
10. 
    
11.     [B,BINT,R] = regress(Y,X) returns a vector R of residuals.
    
12. 
    
13.     [B,BINT,R,RINT] = regress(Y,X) returns a matrix RINT of intervals that
    
14.     can be used to diagnose outliers.  If RINT(i,:) does not contain zero,
    
15.     then the i-th residual is larger than would be expected, at the 5%
    
16.     significance level.  This is evidence that the I-th observation is an
    
17.     outlier.
    
18. 
    
19.     [B,BINT,R,RINT,STATS] = regress(Y,X) returns a vector STATS containing, in
    
20.     the following order, the R-square statistic, the F statistic and p value
    
21.     for the full model, and an estimate of the error variance.
    
22. 
    
23.     [...] = regress(Y,X,ALPHA) uses a 100*(1-ALPHA)% confidence level to
    
24.     compute BINT, and a (100*ALPHA)% significance level to compute RINT.
    
25. 
    
26.     X should include a column of ones so that the model contains a constant
    
27.     term.  The F statistic and p value are computed under the assumption
    
28.     that the model contains a constant term, and they are not correct for
    
29.     models without a constant.  The R-square value is one minus the ratio of
    
30.     the error sum of squares to the total sum of squares.  This value can
    
31.     be negative for models without a constant, which indicates that the
    
32.     model is not appropriate for the data.
    
33. 
    
34.     If columns of X are linearly dependent, regress sets the maximum
    
35.     possible number of elements of B to zero to obtain a "basic solution",
    
36.     and returns zeros in elements of BINT corresponding to the zero
    
37.     elements of B.
    
38. 
    
39.     regress treats NaNs in X or Y as missing values, and removes them.

regress Multiple linear regression using least squares.<br />     B = regress(Y,X) returns the vector B of regression coefficients in the<br />     linear model Y = X*B.  X is an n-by-p design matrix, with rows<br />     corresponding to observations and columns to predictor variables.  Y is<br />     an n-by-1 vector of response observations.<br /> <br />     [B,BINT] = regress(Y,X) returns a matrix BINT of 95% confidence<br />     intervals for B.<br /> <br />     [B,BINT,R] = regress(Y,X) returns a vector R of residuals.<br /> <br />     [B,BINT,R,RINT] = regress(Y,X) returns a matrix RINT of intervals that<br />     can be used to diagnose outliers.  If RINT(i,:) does not contain zero,<br />     then the i-th residual is larger than would be expected, at the 5%<br />     significance level.  This is evidence that the I-th observation is an<br />     outlier.<br /> <br />     [B,BINT,R,RINT,STATS] = regress(Y,X) returns a vector STATS containing, in<br />     the following order, the R-square statistic, the F statistic and p value<br />     for the full model, and an estimate of the error variance.<br /> <br />     [...] = regress(Y,X,ALPHA) uses a 100*(1-ALPHA)% confidence level to<br />     compute BINT, and a (100*ALPHA)% significance level to compute RINT.<br /> <br />     X should include a column of ones so that the model contains a constant<br />     term.  The F statistic and p value are computed under the assumption<br />     that the model contains a constant term, and they are not correct for<br />     models without a constant.  The R-square value is one minus the ratio of<br />     the error sum of squares to the total sum of squares.  This value can<br />     be negative for models without a constant, which indicates that the<br />     model is not appropriate for the data.<br /> <br />     If columns of X are linearly dependent, regress sets the maximum<br />     possible number of elements of B to zero to obtain a "basic solution",<br />     and returns zeros in elements of BINT corresponding to the zero<br />     elements of B.<br /> <br />     regress treats NaNs in X or Y as missing values, and removes them.

秋leaves

liwenhui 发表于 2012-8-27 19:58
使用函数，regress()，具体调用方式见matlab的帮助。

regress Multiple linear regression using least  ...

果断有道翻译了一下

傻人招

数据输入：x1=[1000 600 1200 500 300 400 1300 1100 1300 300]; x2=[5 7 6 6 8 7 5 4 3 9]; y=[100 75 80 70 50 65 90 100 110 60];x=[x1' x2'];回归、检验、预测：rstool(x,y,'purequadratic')   之后得到一个交互画面，给出两幅图形，左边是x2固定时的曲线y(x1)及其置信区间，右边是x1固定时的曲线y(x2)及其置信区间。假如X1为平均收入，x2为价格，y为商品需求量，改变X1或者X2，画面左边的‘predicted y’下方的数据就为对应的y值。  在画面左下方的下拉式菜单中选‘all’，则beta、rmse和residuals都传送到matlab工作区中，在matlab工作区中输入命令 beta，rmse就能得到结果了·~

蓝枫心叶

呵呵，要是楼上的把它变成图文形式就更好了。。。