A machine learning approach for efficient uncertainty quantification using multiscale methods
- Submitting institution
-
Heriot-Watt University
(joint submission with University of Edinburgh)
- Unit of assessment
- 12 - Engineering
- Output identifier
- 16126288
- Type
- D - Journal article
- DOI
-
10.1016/j.jcp.2017.10.034
- Title of journal
- Journal of Computational Physics
- Article number
- -
- First page
- 493
- Volume
- 354
- Issue
- -
- ISSN
- 0021-9991
- Open access status
- Compliant
- Month of publication
- November
- Year of publication
- 2017
- URL
-
-
- Supplementary information
-
-
- Request cross-referral to
- -
- Output has been delayed by COVID-19
- No
- COVID-19 affected output statement
- -
- Forensic science
- No
- Criminology
- No
- Interdisciplinary
- Yes
- Number of additional authors
-
1
- Research group(s)
-
-
- Proposed double-weighted
- No
- Reserve for an output with double weighting
- No
- Additional information
- This work is the one of the first on embedding advanced machine learning algorithms within engineering simulation codes resulting in significant reduction in computing time through leveraging high performance GPUs and TPUs. The method demonstrates the power of multi-scale methods when combined with direct evaluation of sub-scale details using advanced machine learning techniques. This work has been taken up for poroelasticity problems at the Texas A&M University. [Contact available]
- Author contribution statement
- -
- Non-English
- No
- English abstract
- -
