Matches in SemOpenAlex for { <https://semopenalex.org/work/W4220838549> ?p ?o ?g. }
Showing items 1 to 87 of
87
with 100 items per page.
- W4220838549 abstract "Arguably the most popular application task in artificial intelligence is image classification using transfer learning. Transfer learning enables models pre-trained on general classes of images, available in large numbers, to be refined for a specific application. This enables domain experts with their own—generally, substantially smaller—collections of images to build deep learning models. The good performance of such models poses the question of whether it is possible to further reduce the effort required to label training data by adopting a human-in-the-loop interface that presents the expert with the current predictions of the model on a new batch of data and only requires correction of these predictions—rather than de novo labelling by the expert—before retraining the model on the extended data. This paper looks at how to order the data in this iterative training scheme to achieve the highest model performance while minimising the effort needed to correct misclassified examples. Experiments are conducted involving five methods of ordering, using four image classification datasets, and three popular pre-trained models. Two of the methods we consider order the examples a priori whereas the other three employ an active learning approach where the ordering is updated iteratively after each new batch of data and retraining of the model. The main finding is that it is important to consider accuracy of the model in relation to the number of corrections that are required: using accuracy in relation to the number of labelled training examples—as is common practice in the literature—can be misleading. More specifically, active methods require more cumulative corrections than a priori methods for a given level of accuracy. Within their groups, active and a priori methods perform similarly. Preliminary evidence is provided that suggests that for “simple” problems, i.e., those involving fewer examples and classes, no method improves upon random selection of examples. For more complex problems, an a priori strategy based on a greedy sample selection method known as “kernel herding” performs best." @default.
- W4220838549 created "2022-04-03" @default.
- W4220838549 creator A5049127500 @default.
- W4220838549 creator A5059992863 @default.
- W4220838549 creator A5072643673 @default.
- W4220838549 creator A5080148348 @default.
- W4220838549 date "2022-03-22" @default.
- W4220838549 modified "2023-09-28" @default.
- W4220838549 title "Efficiently correcting machine learning: considering the role of example ordering in human-in-the-loop training of image classification models" @default.
- W4220838549 cites W1513874326 @default.
- W4220838549 cites W2119885577 @default.
- W4220838549 cites W2194775991 @default.
- W4220838549 cites W2282821441 @default.
- W4220838549 cites W2977942577 @default.
- W4220838549 cites W2984353870 @default.
- W4220838549 cites W3020996329 @default.
- W4220838549 cites W3036319923 @default.
- W4220838549 cites W3150811281 @default.
- W4220838549 cites W4249817346 @default.
- W4220838549 doi "https://doi.org/10.1145/3490099.3511110" @default.
- W4220838549 hasPublicationYear "2022" @default.
- W4220838549 type Work @default.
- W4220838549 citedByCount "1" @default.
- W4220838549 countsByYear W42208385492023 @default.
- W4220838549 crossrefType "proceedings-article" @default.
- W4220838549 hasAuthorship W4220838549A5049127500 @default.
- W4220838549 hasAuthorship W4220838549A5059992863 @default.
- W4220838549 hasAuthorship W4220838549A5072643673 @default.
- W4220838549 hasAuthorship W4220838549A5080148348 @default.
- W4220838549 hasBestOaLocation W42208385491 @default.
- W4220838549 hasConcept C111472728 @default.
- W4220838549 hasConcept C115961682 @default.
- W4220838549 hasConcept C119857082 @default.
- W4220838549 hasConcept C124101348 @default.
- W4220838549 hasConcept C134306372 @default.
- W4220838549 hasConcept C138885662 @default.
- W4220838549 hasConcept C144133560 @default.
- W4220838549 hasConcept C150899416 @default.
- W4220838549 hasConcept C154945302 @default.
- W4220838549 hasConcept C155202549 @default.
- W4220838549 hasConcept C162324750 @default.
- W4220838549 hasConcept C187736073 @default.
- W4220838549 hasConcept C25343380 @default.
- W4220838549 hasConcept C2778712577 @default.
- W4220838549 hasConcept C2780451532 @default.
- W4220838549 hasConcept C33923547 @default.
- W4220838549 hasConcept C36503486 @default.
- W4220838549 hasConcept C41008148 @default.
- W4220838549 hasConcept C75294576 @default.
- W4220838549 hasConcept C75553542 @default.
- W4220838549 hasConceptScore W4220838549C111472728 @default.
- W4220838549 hasConceptScore W4220838549C115961682 @default.
- W4220838549 hasConceptScore W4220838549C119857082 @default.
- W4220838549 hasConceptScore W4220838549C124101348 @default.
- W4220838549 hasConceptScore W4220838549C134306372 @default.
- W4220838549 hasConceptScore W4220838549C138885662 @default.
- W4220838549 hasConceptScore W4220838549C144133560 @default.
- W4220838549 hasConceptScore W4220838549C150899416 @default.
- W4220838549 hasConceptScore W4220838549C154945302 @default.
- W4220838549 hasConceptScore W4220838549C155202549 @default.
- W4220838549 hasConceptScore W4220838549C162324750 @default.
- W4220838549 hasConceptScore W4220838549C187736073 @default.
- W4220838549 hasConceptScore W4220838549C25343380 @default.
- W4220838549 hasConceptScore W4220838549C2778712577 @default.
- W4220838549 hasConceptScore W4220838549C2780451532 @default.
- W4220838549 hasConceptScore W4220838549C33923547 @default.
- W4220838549 hasConceptScore W4220838549C36503486 @default.
- W4220838549 hasConceptScore W4220838549C41008148 @default.
- W4220838549 hasConceptScore W4220838549C75294576 @default.
- W4220838549 hasConceptScore W4220838549C75553542 @default.
- W4220838549 hasLocation W42208385491 @default.
- W4220838549 hasLocation W42208385492 @default.
- W4220838549 hasOpenAccess W4220838549 @default.
- W4220838549 hasPrimaryLocation W42208385491 @default.
- W4220838549 hasRelatedWork W2772371429 @default.
- W4220838549 hasRelatedWork W2909611106 @default.
- W4220838549 hasRelatedWork W3012393889 @default.
- W4220838549 hasRelatedWork W3138634112 @default.
- W4220838549 hasRelatedWork W4220838549 @default.
- W4220838549 hasRelatedWork W4226361678 @default.
- W4220838549 hasRelatedWork W4281382123 @default.
- W4220838549 hasRelatedWork W4308262314 @default.
- W4220838549 hasRelatedWork W4312018023 @default.
- W4220838549 hasRelatedWork W4312538239 @default.
- W4220838549 isParatext "false" @default.
- W4220838549 isRetracted "false" @default.
- W4220838549 workType "article" @default.