var json_ariticle ={"checkRes":"1","blog_title":"\u3010\u8f6c\u8f7d\u3011Chow\u68c0\u9a8c","uid":"1670266203","blog_id":"638e3d5b0100lruf","blog_body":"

\u90b9\u68c0\u9a8c\uff08Chow test\uff09\u662f\u4e00\u79cd\u8ba1\u91cf\u7ecf\u6d4e<\/A>\u68c0\u9a8c\u3002\u5b83\u53ef\u4ee5\u6d4b\u8bd5\u4e24\u4e2a\u4e0d\u540c\u6570\u636e\u7684\u7ebf\u6027\u56de\u5f52<\/A>\u7684\u7cfb\u6570\u662f\u5426\u76f8\u7b49\u3002\u5728\u65f6\u95f4\u5e8f\u5217\u5206\u6790<\/A>\u4e2d\uff0c\u90b9\u68c0\u9a8c\u88ab\u666e\u904d\u5730\u7528\u6765\u6d4b\u8bd5\u7ed3\u6784\u6027\u53d8\u5316\u662f\u4e0d\u662f\u5b58\u5728\u3002\u90b9\u68c0\u9a8c\u662f\u7531\u7ecf\u6d4e\u5b66\u5bb6\u90b9\u81f3\u5e84<\/A>\u521b\u7acb\u7684\u3002<\/P>\n

\u5047\u8bbe\u6211\u4eec\u7684\u6570\u636e\u6a21\u578b\u662f\uff1a<\/P>\n

\u5982\u679c\u6211\u4eec\u628a\u6570\u636e\u5206\u4e3a\u4e24\u7ec4\uff0c\u90a3\u4e48\uff1a<\/P>\n

\u53ca<\/P>\n

\u90b9\u68c0\u9a8c\u5c31\u662f\u65ad\u5b9a\u662f\u5426a1<\/SUB> = a2<\/SUB>\uff0cb1<\/SUB> =\nb2<\/SUB> \u548c c1<\/SUB> = c2<\/SUB>\u3002<\/P>\n

\u5047\u8bbeSC<\/SUB>\u662f\u7ec4\u5408\u6570\u636e\u7684\u6b8b\u5dee<\/A>\u5e73\u65b9\u548c\uff0cS1<\/SUB>\u662f\u7b2c\u4e00\u7ec4\u6570\u636e\u7684\u6b8b\u5dee\u5e73\u65b9\u548c\uff0cS2<\/SUB>\u662f\u7b2c\u4e8c\u7ec4\u6570\u636e\u7684\u6b8b\u5dee\u5e73\u65b9\u548c\u3002N1<\/SUB>\u548cN2<\/SUB>\u5206\u522b\u662f\u6bcf\u4e00\u7ec4\u6570\u636e\u7684\u89c2\u5bdf\u6570\u76ee\uff0ck\u662f\u53c2\u6570\u7684\u603b\u6570\u3002\u90b9\u68c0\u9a8c\u7684\u68c0\u9a8c\u503c\u662f\uff1a<\/P>\n

\u90b9\u68c0\u9a8c\u7684\u68c0\u9a8c\u503c\u5448F-\u5206\u5e03<\/A>\uff0c\u5b83\u7684\u81ea\u7531\u5ea6<\/A>\u4e3ak\u548cN1<\/SUB>\n+ N2<\/SUB> − 2k\u3002<\/P>\n

 <\/P>\n

The Chow test is a statistical<\/A> and\neconometric<\/A> test of\nwhether the coefficients in two linear\nregressions<\/A> on different data sets are equal. The Chow test was\ninvented by economist \nGregory Chow<\/A>. In econometrics, the Chow test is most commonly\nused in time series\nanalysis<\/A> to test for the presence of a structural\nbreak<\/A>. In program\nevaluation<\/A>, the Chow test is often used to determine whether\nthe independent variables have different impacts on different\nsubgroups of the population.<\/P>\n

Suppose that we model our data as .<\/P>\n

If we split our data into two groups, then we have .<\/P>\n

And .<\/P>\n

The null hypothesis<\/A>\nof the Chow test asserts that a1<\/SUB> = a2<\/SUB>,\nb1<\/SUB> = b2<\/SUB>, and c1<\/SUB> =\nc2<\/SUB>.<\/P>\n

Let SC<\/SUB> be the sum of squared residuals<\/A>\nfrom the combined data, S1<\/SUB> be the sum of squared\nresiduals<\/A>\nfrom the first group, and S2<\/SUB> be the sum of squared\nresiduals<\/A>\nfrom the second group. N1<\/SUB> and N2<\/SUB> are the\nnumber of observations in each group and k is the total number of\nparameters (in this case, 3). Then the Chow test statistic is<\/P>","uname":"\u7d2b\u53f6\u7af9","x_cms_flag":"1","guhost":"","blogtitle":"\u5a09\u30e4\u6c49-\u93c2\u7470\u5d15\u9428\u52eb\u5d25\u7039","tag":"\u6742\u8c08","quote":"","class":"\u6742\u8c08","blog_pubdate":"2010-09-29 10:29:21","blog_2008":"\u65e0"};var status = true;