Details
| Original language | English |
|---|---|
| Title of host publication | Ecomas Proceedia UNCECOMP 2023 |
| Publication status | Published - 2023 |
| Event | 5th ECCOMAS Thematic Conference on Uncertainty Quantification in Computational Sciences and Engineering, UNCECOMP 2023 - Athens, Greece Duration: 12 Jun 2023 → 14 Jun 2023 |
Publication series
| Name | International Conference on Uncertainty Quantification in Computational Science and Engineering |
|---|---|
| Volume | 5 |
| ISSN (Print) | 2623-3339 |
Abstract
Stochastic processes are widely adopted in many domains to deal with problems which are stochastic in nature and involve strong nonlinearity, nonstationarity and uncertain system parameters. However, the uncertainties of spectral representation of the underlying stochastic processes have not been adequately acknowledged due to the data problems in practice, for instance, missing data. Therefore, this paper proposes a novel method for uncertainty quantification of spectral representation in the presence of missing data using Bayesian deep learning models. A range of missing levels are tested. An example in stochastic dynamics is employed for illustration.
Keywords
- Evolutionary power spectral density, LSTM, Missing data, Stochastic Variational inference, Uncertainty Quantification
ASJC Scopus subject areas
- Computer Science(all)
- Computational Theory and Mathematics
- Computer Science(all)
- Computer Science Applications
- Mathematics(all)
- Modelling and Simulation
- Mathematics(all)
- Statistics and Probability
- Mathematics(all)
- Control and Optimization
- Mathematics(all)
- Discrete Mathematics and Combinatorics
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
Ecomas Proceedia UNCECOMP 2023. 2023. (International Conference on Uncertainty Quantification in Computational Science and Engineering; Vol. 5).
Research output: Chapter in book/report/conference proceeding › Contribution to book/anthology › Research › peer review
}
TY - CHAP
T1 - Spectral Density Estimation Of Stochastic Processes Under Missing Data And Uncertainty Quantification With Bayesian Deep Learning
AU - Chen, Yu
AU - Patelli, Edoardo
AU - Edwards, Benjamin
AU - Beer, Michael
N1 - Funding Information: This work was supported by the EU Horizon 2020 - Marie Skłodowska-Curie Actions project URBASIS [Project no. 813137];
PY - 2023
Y1 - 2023
N2 - Stochastic processes are widely adopted in many domains to deal with problems which are stochastic in nature and involve strong nonlinearity, nonstationarity and uncertain system parameters. However, the uncertainties of spectral representation of the underlying stochastic processes have not been adequately acknowledged due to the data problems in practice, for instance, missing data. Therefore, this paper proposes a novel method for uncertainty quantification of spectral representation in the presence of missing data using Bayesian deep learning models. A range of missing levels are tested. An example in stochastic dynamics is employed for illustration.
AB - Stochastic processes are widely adopted in many domains to deal with problems which are stochastic in nature and involve strong nonlinearity, nonstationarity and uncertain system parameters. However, the uncertainties of spectral representation of the underlying stochastic processes have not been adequately acknowledged due to the data problems in practice, for instance, missing data. Therefore, this paper proposes a novel method for uncertainty quantification of spectral representation in the presence of missing data using Bayesian deep learning models. A range of missing levels are tested. An example in stochastic dynamics is employed for illustration.
KW - Evolutionary power spectral density
KW - LSTM
KW - Missing data
KW - Stochastic Variational inference
KW - Uncertainty Quantification
UR - http://www.scopus.com/inward/record.url?scp=85175854631&partnerID=8YFLogxK
U2 - 10.7712/120223.10371.19949
DO - 10.7712/120223.10371.19949
M3 - Contribution to book/anthology
AN - SCOPUS:85175854631
T3 - International Conference on Uncertainty Quantification in Computational Science and Engineering
BT - Ecomas Proceedia UNCECOMP 2023
T2 - 5th ECCOMAS Thematic Conference on Uncertainty Quantification in Computational Sciences and Engineering, UNCECOMP 2023
Y2 - 12 June 2023 through 14 June 2023
ER -