摘要:管理学自相关检验学期论文代写termpaper-Speak from 9 related inspection 9.1 the autocorrelation hypothesis-By chapter 2 know regression model is one of the assumption that the conditions,
管理学自相关检验学期论文
代写termpaper-Speak from 9 related inspection
9.1 the autocorrelation hypo
thesisBy chapter 2 know regression model is one of the assumption that the conditions,
Cov (UI, the uj) = E (UI uj) = 0, (I, j T, I j), (9.1)
That is the value of the error term ut in time is unrelated to each other. Says the error term not from ut related. if
Cov (UI, the uj) 0, (I j)
Said the error term from existing ut related.
Since the correlation and says serial correlation. Originally refers to a random variables in time lag between a with the related. Here is mainly refers to the regression model of random errors with a lag of a related ut relationship. The relationship between autocorrelation is also a.
9.2 first-order autocorrelation
Usually assumed that the error term is related linear. Because of the econometric model from the most common form is related to the first order regression form, so below mainly discussed the error term linear regression of the first order forms, namely
Ut = 1 ut-1 + vt (9.2)
Among them is the 1 regression coefficient, vt is a random error. Vt meet usually hypothesis.
Based on ordinary least square method formula, model (9.2) of the estimation formula is 1,
= (=) (9.3)
Among them T is the sample size. If the ut, u t-1 as two variables, then their correlation coefficient is
= (r =) (9.4)
https://www.51lunwen.org/termpaper/ For large sample is clearly a (9.5)
The formula into the generation (9.4) type
We = (9.6)
So for general parameters have = 1, which first order regression form of regression coefficient is equal to the two variables of the correlation coefficient. So the original regression models of the error term first regression ut form (see model (9.2)) and can be expressed as,
Ut = ut-1 + vt. (9.7)
is the value scope of [1, 1]. When 0, says there is a positive autocorrelation ut; When 0, says there are negative ut autocorrelation. When = 0, says ut does not exist autocorrelation. Figure 9.1 a, c, e, are given respectively from related has the positive, negative and not from the relevant related three sequence. To facilitate understanding of time sequence features related to the positive and negative, figure 9.1 b, d, f, are given respectively figure 9.1 a, c, e, variable in order to one variable lag the scatterplot chart. The positive and negative correlation between the relevant and since the show more and clear.
A. not from relevant sequence diagram b. Not from related scatterplot chart
C. is the related sequences figure d. Is the related scatterplot chart
E. negative since related sequences figure f. negative since related scatterplot chart
Figure 9.1 time series and the related scatterplot chart
Can prove when regression model of the error of the ut exists a regression order form, Cov (UI, the uj) 0. Similarly also can prove when there are high order regression ut form, still have Cov (UI, the uj) 0.
Note: (1) the economic problems from major represented positive autocorrelation. (2) the relevant occurs more often in the time series data.
Since 9.3 related sources and the consequences
There the error term since related, mainly for several reasons.
(1) model with mathematical form. If the use of the mathematical model and the relationship between variables true not consi
本论文由英语论文网提供整理,提供论文代写,英语论文代写,代写论文,代写英语论文,代写留学生论文,代写英文论文,留学生论文代写相关核心关键词搜索。