Sensitivity of the refined descriptive sampling to the population shape characteristics
Year: 2026
Type: article
Abstract: The objective of this work is to deal with the sensitivity of parameters which characterize the shape of a statistical distribution in the case of independent input variables generated using the Refined Descriptive Sampling (RDS) method. We propose algorithms for generating samples using the RDS method according to a given distribution, for estimating shape parameters and for computing Sobol indices. We then estimate the shape parameters for different g -and- k distributions. Furthermore, using linear regression, we have studied the relationship between the estimates of the shape parameters and the g -and- k distribution parameters, finding a good linear approximation. Finally, we investigate the sensitivity analysis using Sobol’s indices. The results have shown that the RDS method provides consistent estimators and is not sensitive for both shape parameters.
Authors: Omar Timeridjine, Megdouda Ourbih-Tari
Institution: University of Béjaïa
Language: English
Subfield: Computational Mechanics
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $2,712
OpenAlex ID: W7235187983
