Spectral Tensor-Train Decomposition for low-rank surrogate models

Daniele Bigoni; Allan Peter Engsig-Karup; Youssef M.  Marzouk

Spectral Tensor-Train Decomposition for low-rank surrogate models

Daniele Bigoni, Allan Peter Engsig-Karup, Youssef M. Marzouk

Massachusetts Institute of Technology

Research output: Contribution to conference › Poster › Research

161 Downloads (Pure)

Abstract

The construction of surrogate models is very important as a mean of acceleration in computational methods for uncertainty quantification (UQ). When the forward model is particularly expensive compared to the accuracy loss due to the use of a surrogate – as for example in computational fluid dynamics (CFD) – the latter can be used for the forward propagation of uncertainty [7] and the solution of inference problems.

Original language	English
Publication date	2014
Number of pages	1
Publication status	Published - 2014
Event	Spatial Statistics and Uncertainty Quantification on Supercomputers - Bath, United Kingdom Duration: 19 May 2014 → 21 May 2014 http://icms.org.uk/workshops/NAISBath

Workshop

Workshop	Spatial Statistics and Uncertainty Quantification on Supercomputers
Country/Territory	United Kingdom
City	Bath
Period	19/05/2014 → 21/05/2014
Internet address	http://icms.org.uk/workshops/NAISBath

Access to Document

PosterBath2014Final published version, 1.83 MB

OpenUrl availability

Full text

Cite this

@conference{4c94d968b209487198b1218bdea7f415,

title = "Spectral Tensor-Train Decomposition for low-rank surrogate models",

abstract = "The construction of surrogate models is very important as a mean of acceleration in computational methods for uncertainty quantification (UQ). When the forward model is particularly expensive compared to the accuracy loss due to the use of a surrogate – as for example in computational fluid dynamics (CFD) – the latter can be used for the forward propagation of uncertainty [7] and the solution of inference problems.",

author = "Daniele Bigoni and Engsig-Karup, {Allan Peter} and Marzouk, {Youssef M.}",

year = "2014",

language = "English",

note = "Spatial Statistics and Uncertainty Quantification on Supercomputers ; Conference date: 19-05-2014 Through 21-05-2014",

url = "http://icms.org.uk/workshops/NAISBath",

}

TY - CONF

T1 - Spectral Tensor-Train Decomposition for low-rank surrogate models

AU - Bigoni, Daniele

AU - Engsig-Karup, Allan Peter

AU - Marzouk, Youssef M.

PY - 2014

Y1 - 2014

N2 - The construction of surrogate models is very important as a mean of acceleration in computational methods for uncertainty quantification (UQ). When the forward model is particularly expensive compared to the accuracy loss due to the use of a surrogate – as for example in computational fluid dynamics (CFD) – the latter can be used for the forward propagation of uncertainty [7] and the solution of inference problems.

AB - The construction of surrogate models is very important as a mean of acceleration in computational methods for uncertainty quantification (UQ). When the forward model is particularly expensive compared to the accuracy loss due to the use of a surrogate – as for example in computational fluid dynamics (CFD) – the latter can be used for the forward propagation of uncertainty [7] and the solution of inference problems.

M3 - Poster

T2 - Spatial Statistics and Uncertainty Quantification on Supercomputers

Y2 - 19 May 2014 through 21 May 2014

ER -

Spectral Tensor-Train Decomposition for low-rank surrogate models

Abstract

Workshop

Access to Document

OpenUrl availability

Fingerprint

Cite this