splus mgarch 迭代过程输出提示一问

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

比如以下程序片段：
ab.bekk1 = mgarch(series = ab, model = bekk1.mod)
......
Iteration  19  Step Size = 0.530913  Likelihood =  5.79470
Iteration  19  Step Size =  1.06183  Likelihood =  5.79459


Convergence R-Square = 0.0000423513 is less than tolerance = 0.0001
Convergence reached.


###################################################

ab.bekk2 = mgarch(series = ab, model = bekk2.mod)
......
Iteration  16  Step Size = 0.000570983  Likelihood =  5.38148
Iteration  16  Step Size = 0.00114197  Likelihood =  5.38148
Iteration  16  Step Size = 0.00228393  Likelihood =  5.38148

Convergence R-Square = 0.001855461
Function value convergence.



请问：
a) Convergence R-Square = 0.0000423513 is less than tolerance = 0.0001
Convergence reached.


b) Convergence R-Square = 0.001855461
Function value convergence.

a得出的结果可靠，还是b得出的可靠？
他们分别说明了什么？

看字面意思，应该是a达到了收敛。但是同样的问题（这里用的LR检验）用wald检验得出的结果却与b符合，与a矛盾。
这里wald检验的结果与经验一致。

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：MGARCH GARCH PLUS ARCH Plu splus splus 提示 提示 输出 输出 MGARCH MGARCH

我在 RATS 的userguide里找到了线索

RATS does allow you to set the optimizers so that they consider convergence achieved if the change in the function value is small. This is done using the instruction NLPAR with the CRTITERION=VALUE option. Convergence on function value is usually quite a bit easier to achieve than convergence on coefficients. You should probably choose this only when just the optimized function value matters, and the cofficients themselves are of minior importance.