Global sensitivity analysis using polynomial chaos expansions
Citations
1,296 citations
1,112 citations
Cites background or methods from "Global sensitivity analysis using p..."
...It is worth mentioning that an algorithm has been devised in [12] to select a minimum number of roots of orthogonal polynomials in the design....
[...]
...Two approaches based on a PC representation of the response are usually distinguished: – the projection approach: each PC coefficient is recast as a multidimensional integral [7, 8] which can be computed either by simulation or quadrature; – the regression approach [9, 10, 11, 12]: the PC coefficients are estimated by minimizing the mean square error of the response approximation....
[...]
...It is also possible to derive inexpensively the global sensitivity indices of the response to the input variables [12]....
[...]
888 citations
Cites methods from "Global sensitivity analysis using p..."
...For example Sudret (2008) presents an approach where generalized polynomial chaos expansions (PCE) are used as emulators and variance-based sensitivity indices (see Section 3.5) are computed analytically as a post-processing of the PCE coefficients....
[...]
710 citations
Cites background or methods from "Global sensitivity analysis using p..."
...Two approaches based on a PC representation of the response are usually distinguished: – the projection approach: each PC coefficient is recast as a multidimensional integral [6, 7] which can be computed either by simulation or quadrature; – the regression approach [8, 9, 10, 11]: the PC coefficients are estimated by minimizing the mean square error of the response approximation in the L2 sense....
[...]
...Indeed, the iterative procedure outlined herein help improve the PC-based GSA method detailed in [11], as shown in [38]....
[...]
...It is however our belief that in most applications, the number of significant terms in the PC expansion is relatively small, because of the two following points: – high order interaction effects are usually negligible compared to main effects and low order interaction effects (this property is referred to as a low effective dimension in applications); – the input variables might have a different impact on the model response, as shown from global sensitivity analysis [11]....
[...]
...It was shown in [10, 11] that a PC approximation of degree p = 2 usually provides satisfactory results for moment and sensitivity analyses, whereas a degree p = 3 is often necessary when performing a reliability analysis....
[...]
694 citations
References
5,495 citations
"Global sensitivity analysis using p..." refers background or methods in this paper
...1 Introduction Polynomial chaos expansions developed back in the 30’s by Wiener (1938) have been brought back in the field of engineering mechanics by Ghanem and Spanos (1991). The usual Hermite chaos is made of the multivariate Hermite polynomials {Ψj, j ∈N} in standard normal variables {ξn, n ∈ N} (called basic variables in the sequel) and allows to represent any random variable with finite variance as follows:...
[...]
...The so-called Spectral stochastic finite element method, named after the pioneering work by (Ghanem and Spanos, 1991) has been applied to many engineering problems (see a review by Sudret and Der Kiureghian (2000) and many recent papers, among...
[...]
4,473 citations