论文部分内容阅读
我觉得陈敦隆同志的“相关分析基木内容及其应用”一文(登于木刊1960年第3期)的目的在于介紹迴归直线,卽从实驗得到自变量x与应变量y的一组数据: x:x_1:x_2,…,x_n,(1) y:y_1,y_2,…,y_n.求直綫y=a+bx使 Q(a,b)=1/n sum from i=1 to n (yi-a-bxi)~2 (2)达到最小。该女所用的工具涉及了偏微商,而严格地說,条件只是Q(a,b)的局部极大与极小所滿足的必要条件,而要說明由(3)解出的(a,b)确实使Q(a,b)达到最小,还需略費小事。本文的目的在于指出一个完全初等的方法来定出(a,b),致使Q(a,b)达到最小。将Q(a,b)展开,并凑平方,得
I think that Comrade Chen Dunlong’s article “Relevant analysis of basic content and its application” (as published in the 3rd issue of the Journal of the 1960) aims to introduce a regression line and obtain a set of independent variable x and dependent variable y from the experiment. Data: x:x_1:x_2,...,x_n,(1) y:y_1,y_2,...,y_n. Find the line y=a+bx so that Q(a,b)=1/n sum from i=1 to n (yi-a-bxi)~2 (2) reaches the minimum. The tools used by the woman involve partial derivatives, and strictly speaking, the conditions are only the necessary conditions for the local maximum and minimum of Q(a,b), and the solutions to the solutions of (a), b) It is really necessary to minimize Q(a,b), and it takes a small amount of work. The purpose of this paper is to point out a completely elementary method to determine (a,b) so that Q(a,b) is minimized. Expand Q(a,b) and square it to get