问题如下图：
选项：
A. 
B. 
C. 
解释：
普通情况下我们用什么修正serial correlation？是加一个AR（2）吗？这里first difference不懂为什么出现。
NO.PZ201709270100000509 问题如下 9.Baseon Exhibit 5, whisingle time-series mol woulmost likely appropriate for Busse to use in precting the future stopriof Company #3? A.Log-linetrenmol B.First-fferenceAR(2) mol C.First-fferencelog AR(1) mol C is correct. a result of the exponentitrenin the time series of stoprices for Company #3, Busse woulwant to take the naturlog of the series anthen first-fferenit. Because the time series also hsericorrelation in the resials from the trenmol, Busse shouluse a more complex mol, suautoregressive (AR) mol. 能不能详细解答下这题是在考什么？为什么最后要用LOG ar
NO.PZ201709270100000509问题如下9.Baseon Exhibit 5, whisingle time-series mol woulmost likely appropriate for Busse to use in precting the future stopriof Company #3? A.Log-linetrenmolB.First-fferenceAR(2) molC.First-fferencelog AR(1) molC is correct. a result of the exponentitrenin the time series of stoprices for Company #3, Busse woulwant to take the naturlog of the series anthen first-fferenit. Because the time series also hsericorrelation in the resials from the trenmol, Busse shouluse a more complex mol, suautoregressive (AR) mol. 一阶差分跟一阶自回归分别用来修正什么问题
NO.PZ201709270100000509 1、First-fferencelog AR(2) mol是不是更加准确？ 2、如果company 3的AR是Yes，那么怎么做？
NO.PZ201709270100000509 老师好， 在之前学的普通的回归分析中，自相关就是残差项自己和自己有一定的关系；条件异方差是残差的取值随着X的波动而波动。 当来到AR模型这块的时候，我发现其实不论是自相关还是条件异方差，都是残差和自己有关系，然后我就不知道怎么区分了。是不是AR模型中的条件异方差是当前的残差和前一个残差的关系，而AR模型中的异方差就不一定了，有可能是当前的残差和前一个残差的关系，也有可能是当前的残差和前几个残差的关系。 请问老师我这么理解对吗？谢谢
First-fferenceAR(2) mol First-fferencelog AR(1) mol C is correct. a result of the exponentitrenin the time series of stoprices for Company #3, Busse woulwant to take the naturlog of the series anthen first-fferenit. Because the time series also hsericorrelation in the resials from the trenmol, Busse shouluse a more complex mol, suautoregressive (AR) mol. 想问一下，课件里哪里讲到如果时间序列有sericorrelation就用AR解决啊？
