x*******u 发帖数: 500 | 1 如果目的是predict, 请问linear regression model是否还需要满足variance
normality和constant variance assumption?
谢谢 |
n*****1 发帖数: 172 | 2 strictly speaking, you don't need normal errors. As long as your sample size
is large and the product of Xe
satisfies some CLT conditions, you still get asymptotic normality.
if you have heteroskedasticity and if you do OLS, then your estimated
coefficients are still unbiased, generally
consistent, but it is not efficient. This means, in EXPECTATION, you get the
right coefficient, but if you want a
confidence interval, then your estimated "range" is not correct.
【在 x*******u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 如果目的是predict, 请问linear regression model是否还需要满足variance : normality和constant variance assumption? : 谢谢
|
x*******u 发帖数: 500 | 3 OK. Since I have large sample size, I do not need normal errors. But if the
plot of residual vs predicted is obvious a megaphone, can I ignore it too?
Thank you. baozi sent. |
D*****a 发帖数: 2847 | 4 You better take care of it.
The SE's will be too small if you use OLS
and the inference can be wrong.
the
【在 x*******u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : OK. Since I have large sample size, I do not need normal errors. But if the : plot of residual vs predicted is obvious a megaphone, can I ignore it too? : Thank you. baozi sent.
|
x*******u 发帖数: 500 | 5 Thank you. baozi sent.
【在 D*****a 的大作中提到】![](/moin_static193/solenoid/img/up.png) : You better take care of it. : The SE's will be too small if you use OLS : and the inference can be wrong. : : the
|
s*r 发帖数: 2757 | 6 predict也有point estimation 和confidence interval estimation的问题呀
【在 x*******u 的大作中提到】![](/moin_static193/solenoid/img/up.png) : 如果目的是predict, 请问linear regression model是否还需要满足variance : normality和constant variance assumption? : 谢谢
|