Matches in SemOpenAlex for { <https://semopenalex.org/work/W3176603669> ?p ?o ?g. }
- W3176603669 abstract "Machine Reading Comprehension (MRC) is an active field in natural language processing with many successful developed models in recent years. Despite their high in-distribution accuracy, these models suffer from two issues: high training cost and low out-of-distribution accuracy. Even though some approaches have been presented to tackle the generalization problem, they have high, intolerable training costs. In this paper, we investigate the effect of ensemble learning approach to improve generalization of MRC systems without retraining a big model. After separately training the base models with different structures on different datasets, they are ensembled using weighting and stacking approaches in probabilistic and non-probabilistic settings. Three configurations are investigated including heterogeneous, homogeneous, and hybrid on eight datasets and six state-of-the-art models. We identify the important factors in the effectiveness of ensemble methods. Also, we compare the robustness of ensemble and fine-tuned models against data distribution shifts. The experimental results show the effectiveness and robustness of the ensemble approach in improving the out-of-distribution accuracy of MRC systems, especially when the base models are similar in accuracies." @default.
- W3176603669 created "2021-07-05" @default.
- W3176603669 creator A5024048254 @default.
- W3176603669 creator A5048602570 @default.
- W3176603669 date "2021-07-01" @default.
- W3176603669 modified "2023-09-27" @default.
- W3176603669 title "Ensemble Learning-Based Approach for Improving Generalization Capability of Machine Reading Comprehension Systems." @default.
- W3176603669 cites W1495683013 @default.
- W3176603669 cites W1544827683 @default.
- W3176603669 cites W1836909716 @default.
- W3176603669 cites W1980264541 @default.
- W3176603669 cites W2023294425 @default.
- W3176603669 cites W2113599721 @default.
- W3176603669 cites W2116064496 @default.
- W3176603669 cites W2150446468 @default.
- W3176603669 cites W2150884987 @default.
- W3176603669 cites W2551396370 @default.
- W3176603669 cites W2557764419 @default.
- W3176603669 cites W2606964149 @default.
- W3176603669 cites W2741239878 @default.
- W3176603669 cites W2746097825 @default.
- W3176603669 cites W2789758093 @default.
- W3176603669 cites W2891691791 @default.
- W3176603669 cites W2892280852 @default.
- W3176603669 cites W2906441821 @default.
- W3176603669 cites W2919420119 @default.
- W3176603669 cites W2951544399 @default.
- W3176603669 cites W2951682790 @default.
- W3176603669 cites W2951873305 @default.
- W3176603669 cites W2962718483 @default.
- W3176603669 cites W2962808855 @default.
- W3176603669 cites W2962881743 @default.
- W3176603669 cites W2962922117 @default.
- W3176603669 cites W2963339397 @default.
- W3176603669 cites W2963341956 @default.
- W3176603669 cites W2963393262 @default.
- W3176603669 cites W2963403868 @default.
- W3176603669 cites W2963515589 @default.
- W3176603669 cites W2963564796 @default.
- W3176603669 cites W2963809228 @default.
- W3176603669 cites W2963863909 @default.
- W3176603669 cites W2963928014 @default.
- W3176603669 cites W2964121744 @default.
- W3176603669 cites W2970019270 @default.
- W3176603669 cites W2970597249 @default.
- W3176603669 cites W2985436116 @default.
- W3176603669 cites W2986836624 @default.
- W3176603669 cites W2988421999 @default.
- W3176603669 cites W2996991341 @default.
- W3176603669 cites W2998369048 @default.
- W3176603669 cites W2998665041 @default.
- W3176603669 cites W3005249048 @default.
- W3176603669 cites W3035267217 @default.
- W3176603669 cites W3089573765 @default.
- W3176603669 cites W3103111734 @default.
- W3176603669 cites W3103417625 @default.
- W3176603669 cites W3161471739 @default.
- W3176603669 hasPublicationYear "2021" @default.
- W3176603669 type Work @default.
- W3176603669 sameAs 3176603669 @default.
- W3176603669 citedByCount "0" @default.
- W3176603669 crossrefType "posted-content" @default.
- W3176603669 hasAuthorship W3176603669A5024048254 @default.
- W3176603669 hasAuthorship W3176603669A5048602570 @default.
- W3176603669 hasConcept C104317684 @default.
- W3176603669 hasConcept C119857082 @default.
- W3176603669 hasConcept C119898033 @default.
- W3176603669 hasConcept C126838900 @default.
- W3176603669 hasConcept C134306372 @default.
- W3176603669 hasConcept C154945302 @default.
- W3176603669 hasConcept C177148314 @default.
- W3176603669 hasConcept C183115368 @default.
- W3176603669 hasConcept C185592680 @default.
- W3176603669 hasConcept C33923547 @default.
- W3176603669 hasConcept C41008148 @default.
- W3176603669 hasConcept C45942800 @default.
- W3176603669 hasConcept C49937458 @default.
- W3176603669 hasConcept C55493867 @default.
- W3176603669 hasConcept C63479239 @default.
- W3176603669 hasConcept C71924100 @default.
- W3176603669 hasConceptScore W3176603669C104317684 @default.
- W3176603669 hasConceptScore W3176603669C119857082 @default.
- W3176603669 hasConceptScore W3176603669C119898033 @default.
- W3176603669 hasConceptScore W3176603669C126838900 @default.
- W3176603669 hasConceptScore W3176603669C134306372 @default.
- W3176603669 hasConceptScore W3176603669C154945302 @default.
- W3176603669 hasConceptScore W3176603669C177148314 @default.
- W3176603669 hasConceptScore W3176603669C183115368 @default.
- W3176603669 hasConceptScore W3176603669C185592680 @default.
- W3176603669 hasConceptScore W3176603669C33923547 @default.
- W3176603669 hasConceptScore W3176603669C41008148 @default.
- W3176603669 hasConceptScore W3176603669C45942800 @default.
- W3176603669 hasConceptScore W3176603669C49937458 @default.
- W3176603669 hasConceptScore W3176603669C55493867 @default.
- W3176603669 hasConceptScore W3176603669C63479239 @default.
- W3176603669 hasConceptScore W3176603669C71924100 @default.
- W3176603669 hasLocation W31766036691 @default.
- W3176603669 hasOpenAccess W3176603669 @default.
- W3176603669 hasPrimaryLocation W31766036691 @default.
- W3176603669 hasRelatedWork W1988664775 @default.