开发者:上海品职教育科技有限公司 隐私政策详情

应用版本:4.2.11(IOS)|3.2.5(安卓)APP下载

Roseline · 2020年07月15日

问一道题:NO.PZ2019052001000137

问题如下:

Even though risk managers cannot eliminate model risk, there are many ways managers can protect themselves against model risk. Which of the following statement about managing model risk is correct?

选项:

A.

Models should be test against known problems.

B.

It is not advisable to estimate model risk using simulations.

C.

Complex models are generally preferable to simple models.

D.

Small discrepancies in model outputs are always acceptable.

解释:

One way to protect against model risk is to test a model against known problems. It is always a good idea to check a model against simple problems to which one already knows the answer, and many problems can be distilled to simple special cases that have known answers. If the model fails to give the correct answer to a problem whose solution is already known, then this indicates that there is something wrong with it.

老师好,能否讲一下选项B,模型风险为什么要用simulation来估计?

1 个答案

品职答疑小助手雍 · 2020年07月16日

b的意思是不建议用,它是错的意味着本意是可以用simulation。

通常是可以通过simulation的过程对模型的参数等等的进行估计,因此通过simulation 的过程对模型风险进行估算也是正常的操作,所以可以使用。

  • 1

    回答
  • 0

    关注
  • 430

    浏览
相关问题

NO.PZ2019052001000137 问题如下 Even though risk managers cannot eliminate mol risk,there are many ways managers cprotethemselves against mol risk. Whichof the following statement about managing mol risk is correct? A.Mols shoultest against known problems. B.It is not aisable to estimate mol risk usingsimulations. C.Complex mols are generally preferable to simplemols. Small screpancies in mol outputs are alwaysacceptable. One wto proteagainst mol risk is to test amol against known problems. It is always a gooia to chea mol againstsimple problems to whione alrea knows the answer, anmany problems cbestilleto simple specicases thhave known answers. If the mol failsto give the correanswer to a problem whose solution is alrea known, thenthis incates ththere is something wrong with it. 老师请问这个知识点在讲义哪里?能否截图一下

2022-11-19 18:20 1 · 回答

NO.PZ2019052001000137 请问什么错?

2021-09-20 14:11 2 · 回答

NO.PZ2019052001000137 麻烦解析下b,我不太懂

2021-09-05 22:15 1 · 回答

NO.PZ2019052001000137 A的话,我记得不是说一个模型好坏,不可以用已知问题来test嘛?因为你这个模型本来就是已知问题作为输入量建立出来的,用已知去测试那肯定是只会得出高度拟合的结论啊?

2021-04-28 23:38 1 · 回答

It is not aisable to estimate mol risk usingsimulations. Complex mols are generally preferable to simplemols. Small screpancies in mol outputs are alwaysacceptable. One wto proteagainst mol risk is to test amol against known problems. It is always a gooia to chea mol againstsimple problems to whione alrea knows the answer, anmany problems cbestilleto simple specicases thhave known answers. If the mol failsto give the correanswer to a problem whose solution is alrea known, thenthis incates ththere is something wrong with it. 老师 C 错在哪里 不是鼓励使用简单模型吗

2020-10-04 12:47 1 · 回答