Abstract | ||
---|---|---|
A novel class of models, referred to as polynomial chaos (PC) translation models, is developed for non-Gaussian vectors. The models match target marginal distributions exactly and dependence structures approximately. They are nonlinear transformations of truncated PC expansions whose coefficients are selected to best describe specified quantities of interest. Optimization algorithms are used to implement PC translation models. The computation demands to implement PC translation models and truncated PC expansions are similar. Numerical examples are presented to illustrate the implementation and performance of PC translation models. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1137/18M118061X | SIAM JOURNAL ON SCIENTIFIC COMPUTING |
Keywords | Field | DocType |
dependence,Monte Carlo,polynomial chaos,tail independence,spectral measure | Applied mathematics,Monte Carlo method,Spectral measure,Nonlinear system,Mathematical analysis,Multivariate statistics,Polynomial chaos,Optimization algorithm,Marginal distribution,Mathematics,Computation | Journal |
Volume | Issue | ISSN |
41 | 2 | 1064-8275 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
1 |
Name | Order | Citations | PageRank |
---|---|---|---|
Mircea Grigoriu | 1 | 4 | 3.83 |