Actuarial Outpost
 
Go Back   Actuarial Outpost > Actuarial Discussion Forum > Software & Technology
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions


Fill in a brief DW Simpson Registration Form
to be contacted when new jobs meet your criteria.


Reply
 
Thread Tools Search this Thread Display Modes
  #21  
Old 09-12-2017, 02:30 PM
DiscreteAndDiscreet DiscreteAndDiscreet is offline
Member
AAA
 
Join Date: May 2016
Posts: 478
Default

Sensitivity of a Monte Carlo method to issues with RNG quality are related to the computational difficulty of the underlying problem. I would be very surprised if an actuarial model had materially biased results on a PRNG compared to a hardware RNG. In physics problems where that kind of issue comes up, the problem is that the phase space is not smooth and there are abrupt, large changes in the output and the calculation is sensitive to the uniformity of sampling in these regions. That's very much at odds with the typical characteristics of actuarial models.
Reply With Quote
  #22  
Old 09-12-2017, 03:22 PM
echo echo is offline
Member
 
Join Date: Nov 2002
Posts: 875
Default

Quote:
Originally Posted by DiscreteAndDiscreet View Post
Sensitivity of a Monte Carlo method to issues with RNG quality are related to the computational difficulty of the underlying problem. I would be very surprised if an actuarial model had materially biased results on a PRNG compared to a hardware RNG. In physics problems where that kind of issue comes up, the problem is that the phase space is not smooth and there are abrupt, large changes in the output and the calculation is sensitive to the uniformity of sampling in these regions. That's very much at odds with the typical characteristics of actuarial models.
Great, you're just begging the pricing people to design a product that's sensitive to RNG quality.
Reply With Quote
  #23  
Old 09-12-2017, 05:51 PM
DiscreteAndDiscreet DiscreteAndDiscreet is offline
Member
AAA
 
Join Date: May 2016
Posts: 478
Default

Quote:
Originally Posted by echo View Post
Great, you're just begging the pricing people to design a product that's sensitive to RNG quality.


The RNG itself is a parameter to the model and excessive sensitivity to parameters is generally a strong sign that you don't want to be responsible for picking a price for something. If you're working with people who have trouble recognizing this, I think if you wait long enough natural selection will take its course and they'll fall over the railing at a zoo or wander into an active construction site.

In the meantime, keep a good distance away from anyone who insists on using a model where the results can't easily be estimated from a simpler model.
Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


All times are GMT -4. The time now is 11:12 PM.


Powered by vBulletin®
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.
*PLEASE NOTE: Posts are not checked for accuracy, and do not
represent the views of the Actuarial Outpost or its sponsors.
Page generated in 0.27854 seconds with 11 queries