New papers published in the Journal of Computational Physics and International Journal for Uncertainty Quantification
Two papers about the work on Stochastic Spectral Embedding (SSE) by P.-R. Wagner, S. Marelli, C. Lataniotis and B. Sudret have been recently published.
The first paper, entitled stochastic spectral embedding (SSE) proposes a new way for surrogate modelling by constructing residual spectral expansions in subdomains of the input space. This technique is particularly powerful for approximating models with complex local characteristics. The publication can be found external page here and the associated report on our internal archive here.
The second paper, entitled Bayesian model inversion using stochastic spectral embedding extends the SSE method with an active learning scheme to adaptively enrich the experimental design for approximating the likelihood functions ocurring in Bayesian inverse problems. Visit external page this page for the publication and this page for the report on our internal archive.