Author Topic: QuoFEM Sensitivity Analysis  (Read 12419 times)

pellumbz

  • Newbie
  • *
  • Posts: 9
    • View Profile
QuoFEM Sensitivity Analysis
« on: December 16, 2022, 10:38:39 AM »
Dear all,

I'm running sensitivity analysis in QuoFEM, and I have a question related to "UQ Engine".
In QuoFEM, you can perform sensitivity analysis by using either Dakota or SimCenterUQ  UQ engine.
I run the same case (same RV and number of samples)  with both UQ engines but I don't get the same results regarding the sensitivity of each parameter (main and total).
My question is, which is more accurate or reliable regarding the sensitivity analysis?
Should both UQ Engine provide the same results or is normal that the results are different?
I attached the results from both analysis for your consideration.

BR,
Pellumb

Sang-ri

  • Administrator
  • Jr. Member
  • *****
  • Posts: 70
    • View Profile
Re: QuoFEM Sensitivity Analysis
« Reply #1 on: December 17, 2022, 01:07:16 AM »
Hello Pellumb, thanks for the question and for sharing the results!

It is likely that different algorithms will produce slightly different results because of the (1) sampling variability and (2) different assumptions that each algorithm makes. In this case, because you were able to run enough number of simulations, the results from dakota engine are likely more accurate and thus preferred.

The method in the dakota engine ( efficient monte-carlo ) is asymptotically unbiased, meaning it is guaranteed to converge to the 'exact' values when a large number of samples are available. If you specify 1500 samples, the results should be pretty accurate in most applications.

The approach in SimCenterUQ engine ( PM-GSA ), on the other hand, introduces more assumptions to achieve faster convergence. Because of these assumptions, even when we run enough number of simulations, the results may still be biased. However, there are situations where this method is preferred to the one above, for example: (1) when the simulation model is very expensive, so only a limited number of samples (maybe a few hundred) are available; (2) when the random variables are correlated; (3) when Monte Carlo samples are already available, so you want to directly import the dataset instead of running simulations again; (4) when you would like to calculate 'joint sensitivity indices' or 'higher-order sensitivity indices'

Hope this will help!

Sang-ri

pellumbz

  • Newbie
  • *
  • Posts: 9
    • View Profile
Re: QuoFEM Sensitivity Analysis
« Reply #2 on: December 19, 2022, 08:51:18 AM »
Dear Dr. Sang-ri,
I would like to thank you, I really appreciate your effort. 
Many thanks also for explanation, it is clear for me now :)

Best regards,
Pellumb