Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification
Inverse Uncertainty Quantification (UQ), or Bayesian calibration, is the process to quantify the uncertainties of random input parameters based on experimental data. The introduction of model discrepancy term is significant because “over-fitting” can theoretically be avoided. But it also poses chall...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Elsevier BV
2020
|
| Online Access: | https://hdl.handle.net/1721.1/124474 |
| _version_ | 1826189650907103232 |
|---|---|
| author | Wu, Xu Shirvan, Koroush Kozlowski, Tomasz |
| author2 | Massachusetts Institute of Technology. Department of Nuclear Science and Engineering |
| author_facet | Massachusetts Institute of Technology. Department of Nuclear Science and Engineering Wu, Xu Shirvan, Koroush Kozlowski, Tomasz |
| author_sort | Wu, Xu |
| collection | MIT |
| description | Inverse Uncertainty Quantification (UQ), or Bayesian calibration, is the process to quantify the uncertainties of random input parameters based on experimental data. The introduction of model discrepancy term is significant because “over-fitting” can theoretically be avoided. But it also poses challenges in the practical applications. One of the mostly concerned and unresolved problem is the “lack of identifiability” issue. With the presence of model discrepancy, inverse UQ becomes “non-identifiable” in the sense that it is difficult to precisely distinguish between the parameter uncertainties and model discrepancy when estimating the calibration parameters. Previous research to alleviate the non-identifiability issue focused on using informative priors for the calibration parameters and the model discrepancy, which is usually not a viable solution because one rarely has such accurate and informative prior knowledge. In this work, we show that identifiability is largely related to the sensitivity of the calibration parameters with regards to the chosen responses. We adopted an improved modular Bayesian approach for inverse UQ that does not require priors for the model discrepancy term. The relationship between sensitivity and identifiability was demonstrated with a practical example in nuclear engineering. It was shown that, in order for a certain calibration parameter to be statistically identifiable, it should be significant to at least one of the responses whose data are used for inverse UQ. Good identifiability cannot be achieved for a certain calibration parameter if it is not significant to any of the responses. It is also demonstrated that “fake identifiability” is possible if model responses are not appropriately chosen, or if inaccurate but informative prior distributions are specified. ©2019 |
| first_indexed | 2024-09-23T08:18:59Z |
| format | Article |
| id | mit-1721.1/124474 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T08:18:59Z |
| publishDate | 2020 |
| publisher | Elsevier BV |
| record_format | dspace |
| spelling | mit-1721.1/1244742022-09-23T12:16:56Z Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification Wu, Xu Shirvan, Koroush Kozlowski, Tomasz Massachusetts Institute of Technology. Department of Nuclear Science and Engineering Inverse Uncertainty Quantification (UQ), or Bayesian calibration, is the process to quantify the uncertainties of random input parameters based on experimental data. The introduction of model discrepancy term is significant because “over-fitting” can theoretically be avoided. But it also poses challenges in the practical applications. One of the mostly concerned and unresolved problem is the “lack of identifiability” issue. With the presence of model discrepancy, inverse UQ becomes “non-identifiable” in the sense that it is difficult to precisely distinguish between the parameter uncertainties and model discrepancy when estimating the calibration parameters. Previous research to alleviate the non-identifiability issue focused on using informative priors for the calibration parameters and the model discrepancy, which is usually not a viable solution because one rarely has such accurate and informative prior knowledge. In this work, we show that identifiability is largely related to the sensitivity of the calibration parameters with regards to the chosen responses. We adopted an improved modular Bayesian approach for inverse UQ that does not require priors for the model discrepancy term. The relationship between sensitivity and identifiability was demonstrated with a practical example in nuclear engineering. It was shown that, in order for a certain calibration parameter to be statistically identifiable, it should be significant to at least one of the responses whose data are used for inverse UQ. Good identifiability cannot be achieved for a certain calibration parameter if it is not significant to any of the responses. It is also demonstrated that “fake identifiability” is possible if model responses are not appropriately chosen, or if inaccurate but informative prior distributions are specified. ©2019 2020-04-01T14:12:29Z 2020-04-01T14:12:29Z 2019-11 2018-10 2020-02-27T18:15:40Z Article http://purl.org/eprint/type/JournalArticle 1090-2716 0021-9991 https://hdl.handle.net/1721.1/124474 Wu, Xu, Koroush Shirvan, and Tomasz Kozlowski, "Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification." Journal of computational physics 396, 1 (November 2019): p. 12-30 doi 10.1016/j.jcp.2019.06.032 ©2019 Author(s) Keywords: inverse uncertainty quantification; modular Bayesian approach; identifiability; sensitivity en 10.1016/J.JCP.2019.06.032 Journal of computational physics Creative Commons Attribution-NonCommercial-NoDerivs License http://creativecommons.org/licenses/by-nc-nd/4.0/ application/pdf Elsevier BV arXiv |
| spellingShingle | Wu, Xu Shirvan, Koroush Kozlowski, Tomasz Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| title | Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| title_full | Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| title_fullStr | Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| title_full_unstemmed | Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| title_short | Demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| title_sort | demonstration of the relationship between sensitivity and identifiability for inverse uncertainty quantification |
| url | https://hdl.handle.net/1721.1/124474 |
| work_keys_str_mv | AT wuxu demonstrationoftherelationshipbetweensensitivityandidentifiabilityforinverseuncertaintyquantification AT shirvankoroush demonstrationoftherelationshipbetweensensitivityandidentifiabilityforinverseuncertaintyquantification AT kozlowskitomasz demonstrationoftherelationshipbetweensensitivityandidentifiabilityforinverseuncertaintyquantification |