Matches in SemOpenAlex for { <https://semopenalex.org/work/W3198509990> ?p ?o ?g. }
Showing items 1 to 82 of
82
with 100 items per page.
- W3198509990 abstract "PreviousNext No AccessFirst International Meeting for Applied Geoscience & Energy Expanded AbstractsImproved seismic impedance inversion based on uncertainty analysisAuthors: Qiming MaYuqing WangQi WangWenkai LuQiming MaTsinghua UniversitySearch for more papers by this author, Yuqing WangTsinghua UniversitySearch for more papers by this author, Qi WangTsinghua UniversitySearch for more papers by this author, and Wenkai LuTsinghua UniversitySearch for more papers by this authorhttps://doi.org/10.1190/segam2021-3583441.1 SectionsSupplemental MaterialAboutPDF/ePub ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InRedditEmail AbstractDeep learning methods are widely applied in the geophysical field recently. Nevertheless, most of these methods only give predicted results without uncertainty analysis. In real geological exploration, we only know little well-logging data, not all the ground truth. It is beneficial to use uncertainty to evaluate the model and the inversion impedance. In this abstract, we mainly made two contributions: (1) We proposed an uncertainty analysis framework for impedance inversion that can estimate the impedance and the epistemic uncertainty simultaneously. The epistemic uncertainty can be used to evaluate the generalization ability of the trained model and the inversion impedance directly. (2) We further improved the inversion results by adopting the uncertainty as the weight of the loss function. We accomplish this by placing evidential priors over the original Gaussian likelihood function and training the model to infer the hyper-parameters of the evidential distribution. By adding the uncertainty backpropagation, we improve the inversion accuracy by 17.6%.Keywords: impedance, inversion, machine learningPermalink: https://doi.org/10.1190/segam2021-3583441.1FiguresReferencesRelatedDetails First International Meeting for Applied Geoscience & Energy Expanded AbstractsISSN (print):1052-3812 ISSN (online):1949-4645Copyright: 2021 Pages: 3561 publication data© 2021 Published in electronic format with permission by the Society of Exploration GeophysicistsPublisher:Society of Exploration Geophysicists HistoryPublished: 01 Sep 2021 CITATION INFORMATION Qiming Ma, Yuqing Wang, Qi Wang, and Wenkai Lu, (2021), Improved seismic impedance inversion based on uncertainty analysis, SEG Technical Program Expanded Abstracts : 1400-1404. https://doi.org/10.1190/segam2021-3583441.1 Plain-Language Summary Keywordsimpedanceinversionmachine learningPDF DownloadLoading ..." @default.
- W3198509990 created "2021-09-13" @default.
- W3198509990 creator A5011509629 @default.
- W3198509990 creator A5015195367 @default.
- W3198509990 creator A5016059614 @default.
- W3198509990 creator A5090655358 @default.
- W3198509990 date "2021-09-01" @default.
- W3198509990 modified "2023-09-27" @default.
- W3198509990 title "Improved seismic impedance inversion based on uncertainty analysis" @default.
- W3198509990 cites W2043112821 @default.
- W3198509990 cites W2053376610 @default.
- W3198509990 cites W2967196803 @default.
- W3198509990 cites W2968094316 @default.
- W3198509990 cites W2968528446 @default.
- W3198509990 doi "https://doi.org/10.1190/segam2021-3583441.1" @default.
- W3198509990 hasPublicationYear "2021" @default.
- W3198509990 type Work @default.
- W3198509990 sameAs 3198509990 @default.
- W3198509990 citedByCount "2" @default.
- W3198509990 countsByYear W31985099902022 @default.
- W3198509990 crossrefType "proceedings-article" @default.
- W3198509990 hasAuthorship W3198509990A5011509629 @default.
- W3198509990 hasAuthorship W3198509990A5015195367 @default.
- W3198509990 hasAuthorship W3198509990A5016059614 @default.
- W3198509990 hasAuthorship W3198509990A5090655358 @default.
- W3198509990 hasConcept C107673813 @default.
- W3198509990 hasConcept C11413529 @default.
- W3198509990 hasConcept C119599485 @default.
- W3198509990 hasConcept C119857082 @default.
- W3198509990 hasConcept C127313418 @default.
- W3198509990 hasConcept C127413603 @default.
- W3198509990 hasConcept C134306372 @default.
- W3198509990 hasConcept C154945302 @default.
- W3198509990 hasConcept C165205528 @default.
- W3198509990 hasConcept C177148314 @default.
- W3198509990 hasConcept C177769412 @default.
- W3198509990 hasConcept C177803969 @default.
- W3198509990 hasConcept C17829176 @default.
- W3198509990 hasConcept C1893757 @default.
- W3198509990 hasConcept C32230216 @default.
- W3198509990 hasConcept C33923547 @default.
- W3198509990 hasConcept C41008148 @default.
- W3198509990 hasConcept C44154836 @default.
- W3198509990 hasConcept C77928131 @default.
- W3198509990 hasConcept C8058405 @default.
- W3198509990 hasConceptScore W3198509990C107673813 @default.
- W3198509990 hasConceptScore W3198509990C11413529 @default.
- W3198509990 hasConceptScore W3198509990C119599485 @default.
- W3198509990 hasConceptScore W3198509990C119857082 @default.
- W3198509990 hasConceptScore W3198509990C127313418 @default.
- W3198509990 hasConceptScore W3198509990C127413603 @default.
- W3198509990 hasConceptScore W3198509990C134306372 @default.
- W3198509990 hasConceptScore W3198509990C154945302 @default.
- W3198509990 hasConceptScore W3198509990C165205528 @default.
- W3198509990 hasConceptScore W3198509990C177148314 @default.
- W3198509990 hasConceptScore W3198509990C177769412 @default.
- W3198509990 hasConceptScore W3198509990C177803969 @default.
- W3198509990 hasConceptScore W3198509990C17829176 @default.
- W3198509990 hasConceptScore W3198509990C1893757 @default.
- W3198509990 hasConceptScore W3198509990C32230216 @default.
- W3198509990 hasConceptScore W3198509990C33923547 @default.
- W3198509990 hasConceptScore W3198509990C41008148 @default.
- W3198509990 hasConceptScore W3198509990C44154836 @default.
- W3198509990 hasConceptScore W3198509990C77928131 @default.
- W3198509990 hasConceptScore W3198509990C8058405 @default.
- W3198509990 hasLocation W31985099901 @default.
- W3198509990 hasOpenAccess W3198509990 @default.
- W3198509990 hasPrimaryLocation W31985099901 @default.
- W3198509990 hasRelatedWork W1870960091 @default.
- W3198509990 hasRelatedWork W1995682383 @default.
- W3198509990 hasRelatedWork W2186219592 @default.
- W3198509990 hasRelatedWork W2218480601 @default.
- W3198509990 hasRelatedWork W2417273483 @default.
- W3198509990 hasRelatedWork W3175794431 @default.
- W3198509990 hasRelatedWork W3196516528 @default.
- W3198509990 hasRelatedWork W3198509990 @default.
- W3198509990 hasRelatedWork W3213237859 @default.
- W3198509990 hasRelatedWork W4285272371 @default.
- W3198509990 isParatext "false" @default.
- W3198509990 isRetracted "false" @default.
- W3198509990 magId "3198509990" @default.
- W3198509990 workType "article" @default.