Dakota Reference Manual
Version 6.4
LargeScale Engineering Optimization and Uncertainty Analysis

Allow refinement of probability and generalized reliability results using importance sampling
This keyword is related to the topics:
Alias: sample_refinement
Argument(s): none
Default: no refinement
Required/Optional  Description of Group  Dakota Keyword  Dakota Keyword Description  

Required (Choose One)  Group 1  import  Sampling option  
adapt_import  Importance sampling option  
mm_adapt_import  Sampling option  
Optional  refinement_samples  Number of samples used to refine a probabilty estimate or sampling design. 
The probability_refinement
allows refinement of probability and generalized reliability results using importance sampling. If one specifies probability_refinement
, there are some additional options. One can specify which type of importance sampling to use (import
, adapt_import
, or mm_adapt_import
). Additionally, one can specify the number of refinement samples to use with refinement_samples
and the seed to use with seed
.
The probability_refinement
density reweighting accounts originally was developed based on Gaussian distributions. It now accounts for additional nonGaussian cases.