Matches in SemOpenAlex for { <https://semopenalex.org/work/W2904314507> ?p ?o ?g. }
- W2904314507 abstract "Ensemble learning is a mainstay in modern data science practice. Conventional ensemble algorithms assigns to base models a set of deterministic, constant model weights that (1) do not fully account for variations in base model accuracy across subgroups, nor (2) provide uncertainty estimates for the ensemble prediction, which could result in mis-calibrated (i.e. precise but biased) predictions that could in turn negatively impact the algorithm performance in real-word applications. In this work, we present an adaptive, probabilistic approach to ensemble learning using dependent tail-free process as ensemble weight prior. Given input feature $mathbf{x}$, our method optimally combines base models based on their predictive accuracy in the feature space $mathbf{x} in mathcal{X}$, and provides interpretable uncertainty estimates both in model selection and in ensemble prediction. To encourage scalable and calibrated inference, we derive a structured variational inference algorithm that jointly minimize KL objective and the model's calibration score (i.e. Continuous Ranked Probability Score (CRPS)). We illustrate the utility of our method on both a synthetic nonlinear function regression task, and on the real-world application of spatio-temporal integration of particle pollution prediction models in New England." @default.
- W2904314507 created "2018-12-22" @default.
- W2904314507 creator A5021008013 @default.
- W2904314507 creator A5040162395 @default.
- W2904314507 creator A5045303930 @default.
- W2904314507 creator A5073573091 @default.
- W2904314507 date "2018-12-08" @default.
- W2904314507 modified "2023-09-27" @default.
- W2904314507 title "Adaptive and Calibrated Ensemble Learning with Dependent Tail-free Process." @default.
- W2904314507 cites W137285897 @default.
- W2904314507 cites W1999617766 @default.
- W2904314507 cites W2000109687 @default.
- W2904314507 cites W2001272401 @default.
- W2904314507 cites W2025720061 @default.
- W2904314507 cites W2063608735 @default.
- W2904314507 cites W2121360982 @default.
- W2904314507 cites W2130715829 @default.
- W2904314507 cites W2146766088 @default.
- W2904314507 cites W2153185114 @default.
- W2904314507 cites W2272706126 @default.
- W2904314507 cites W2310114729 @default.
- W2904314507 cites W2509334837 @default.
- W2904314507 cites W2617480188 @default.
- W2904314507 cites W2737684225 @default.
- W2904314507 cites W2810586154 @default.
- W2904314507 cites W2898407312 @default.
- W2904314507 cites W2914369697 @default.
- W2904314507 cites W2951493172 @default.
- W2904314507 cites W3101380508 @default.
- W2904314507 hasPublicationYear "2018" @default.
- W2904314507 type Work @default.
- W2904314507 sameAs 2904314507 @default.
- W2904314507 citedByCount "0" @default.
- W2904314507 crossrefType "posted-content" @default.
- W2904314507 hasAuthorship W2904314507A5021008013 @default.
- W2904314507 hasAuthorship W2904314507A5040162395 @default.
- W2904314507 hasAuthorship W2904314507A5045303930 @default.
- W2904314507 hasAuthorship W2904314507A5073573091 @default.
- W2904314507 hasConcept C105795698 @default.
- W2904314507 hasConcept C11413529 @default.
- W2904314507 hasConcept C119857082 @default.
- W2904314507 hasConcept C119898033 @default.
- W2904314507 hasConcept C134306372 @default.
- W2904314507 hasConcept C138885662 @default.
- W2904314507 hasConcept C154945302 @default.
- W2904314507 hasConcept C165838908 @default.
- W2904314507 hasConcept C2776214188 @default.
- W2904314507 hasConcept C2776401178 @default.
- W2904314507 hasConcept C33923547 @default.
- W2904314507 hasConcept C41008148 @default.
- W2904314507 hasConcept C41895202 @default.
- W2904314507 hasConcept C42058472 @default.
- W2904314507 hasConcept C45942800 @default.
- W2904314507 hasConcept C48044578 @default.
- W2904314507 hasConcept C49937458 @default.
- W2904314507 hasConcept C77088390 @default.
- W2904314507 hasConceptScore W2904314507C105795698 @default.
- W2904314507 hasConceptScore W2904314507C11413529 @default.
- W2904314507 hasConceptScore W2904314507C119857082 @default.
- W2904314507 hasConceptScore W2904314507C119898033 @default.
- W2904314507 hasConceptScore W2904314507C134306372 @default.
- W2904314507 hasConceptScore W2904314507C138885662 @default.
- W2904314507 hasConceptScore W2904314507C154945302 @default.
- W2904314507 hasConceptScore W2904314507C165838908 @default.
- W2904314507 hasConceptScore W2904314507C2776214188 @default.
- W2904314507 hasConceptScore W2904314507C2776401178 @default.
- W2904314507 hasConceptScore W2904314507C33923547 @default.
- W2904314507 hasConceptScore W2904314507C41008148 @default.
- W2904314507 hasConceptScore W2904314507C41895202 @default.
- W2904314507 hasConceptScore W2904314507C42058472 @default.
- W2904314507 hasConceptScore W2904314507C45942800 @default.
- W2904314507 hasConceptScore W2904314507C48044578 @default.
- W2904314507 hasConceptScore W2904314507C49937458 @default.
- W2904314507 hasConceptScore W2904314507C77088390 @default.
- W2904314507 hasLocation W29043145071 @default.
- W2904314507 hasOpenAccess W2904314507 @default.
- W2904314507 hasPrimaryLocation W29043145071 @default.
- W2904314507 hasRelatedWork W1749161648 @default.
- W2904314507 hasRelatedWork W2134650532 @default.
- W2904314507 hasRelatedWork W2157573406 @default.
- W2904314507 hasRelatedWork W2512115115 @default.
- W2904314507 hasRelatedWork W2551342063 @default.
- W2904314507 hasRelatedWork W2592168475 @default.
- W2904314507 hasRelatedWork W2786902984 @default.
- W2904314507 hasRelatedWork W2902010252 @default.
- W2904314507 hasRelatedWork W2909854773 @default.
- W2904314507 hasRelatedWork W2921129588 @default.
- W2904314507 hasRelatedWork W2953263857 @default.
- W2904314507 hasRelatedWork W2966294393 @default.
- W2904314507 hasRelatedWork W3000658058 @default.
- W2904314507 hasRelatedWork W3004570229 @default.
- W2904314507 hasRelatedWork W3009946576 @default.
- W2904314507 hasRelatedWork W3045135582 @default.
- W2904314507 hasRelatedWork W3122793235 @default.
- W2904314507 hasRelatedWork W3136163546 @default.
- W2904314507 hasRelatedWork W3147502964 @default.
- W2904314507 hasRelatedWork W3213893529 @default.
- W2904314507 isParatext "false" @default.
- W2904314507 isRetracted "false" @default.
- W2904314507 magId "2904314507" @default.