Matches in SemOpenAlex for { <https://semopenalex.org/work/W4381586568> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4381586568 abstract "Despite deep learning has achieved great success, it often relies on a large amount of training data with accurate labels, which are expensive and time-consuming to collect. A prominent direction to reduce the cost is to learn with noisy labels, which are ubiquitous in the real-world applications. A critical challenge for such a learning task is to reduce the effect of network memorization on the falsely-labeled data. In this work, we propose an iterative selection approach based on the Weibull mixture model, which identifies clean data by considering the overall learning dynamics of each data instance. In contrast to the previous small-loss heuristics, we leverage the observation that deep network is easy to memorize and hard to forget clean data. In particular, we measure the difficulty of memorization and forgetting for each instance via the transition times between being misclassified and being memorized in training, and integrate them into a novel metric for selection. Based on the proposed metric, we retain a subset of identified clean data and repeat the selection procedure to iteratively refine the clean subset, which is finally used for model training. To validate our method, we perform extensive experiments on synthetic noisy datasets and real-world web data, and our strategy outperforms existing noisy-label learning methods." @default.
- W4381586568 created "2023-06-22" @default.
- W4381586568 creator A5015970030 @default.
- W4381586568 creator A5062881932 @default.
- W4381586568 creator A5074627818 @default.
- W4381586568 creator A5088442800 @default.
- W4381586568 date "2023-06-20" @default.
- W4381586568 modified "2023-10-18" @default.
- W4381586568 title "MILD: Modeling the Instance Learning Dynamics for Learning with Noisy Labels" @default.
- W4381586568 doi "https://doi.org/10.48550/arxiv.2306.11560" @default.
- W4381586568 hasPublicationYear "2023" @default.
- W4381586568 type Work @default.
- W4381586568 citedByCount "0" @default.
- W4381586568 crossrefType "posted-content" @default.
- W4381586568 hasAuthorship W4381586568A5015970030 @default.
- W4381586568 hasAuthorship W4381586568A5062881932 @default.
- W4381586568 hasAuthorship W4381586568A5074627818 @default.
- W4381586568 hasAuthorship W4381586568A5088442800 @default.
- W4381586568 hasBestOaLocation W43815865681 @default.
- W4381586568 hasConcept C108583219 @default.
- W4381586568 hasConcept C111919701 @default.
- W4381586568 hasConcept C119857082 @default.
- W4381586568 hasConcept C124101348 @default.
- W4381586568 hasConcept C127705205 @default.
- W4381586568 hasConcept C138885662 @default.
- W4381586568 hasConcept C145420912 @default.
- W4381586568 hasConcept C153083717 @default.
- W4381586568 hasConcept C154945302 @default.
- W4381586568 hasConcept C162324750 @default.
- W4381586568 hasConcept C176217482 @default.
- W4381586568 hasConcept C187736073 @default.
- W4381586568 hasConcept C21547014 @default.
- W4381586568 hasConcept C2780451532 @default.
- W4381586568 hasConcept C30038468 @default.
- W4381586568 hasConcept C33923547 @default.
- W4381586568 hasConcept C41008148 @default.
- W4381586568 hasConcept C41895202 @default.
- W4381586568 hasConcept C7149132 @default.
- W4381586568 hasConcept C81917197 @default.
- W4381586568 hasConcept C93959086 @default.
- W4381586568 hasConcept C95623464 @default.
- W4381586568 hasConceptScore W4381586568C108583219 @default.
- W4381586568 hasConceptScore W4381586568C111919701 @default.
- W4381586568 hasConceptScore W4381586568C119857082 @default.
- W4381586568 hasConceptScore W4381586568C124101348 @default.
- W4381586568 hasConceptScore W4381586568C127705205 @default.
- W4381586568 hasConceptScore W4381586568C138885662 @default.
- W4381586568 hasConceptScore W4381586568C145420912 @default.
- W4381586568 hasConceptScore W4381586568C153083717 @default.
- W4381586568 hasConceptScore W4381586568C154945302 @default.
- W4381586568 hasConceptScore W4381586568C162324750 @default.
- W4381586568 hasConceptScore W4381586568C176217482 @default.
- W4381586568 hasConceptScore W4381586568C187736073 @default.
- W4381586568 hasConceptScore W4381586568C21547014 @default.
- W4381586568 hasConceptScore W4381586568C2780451532 @default.
- W4381586568 hasConceptScore W4381586568C30038468 @default.
- W4381586568 hasConceptScore W4381586568C33923547 @default.
- W4381586568 hasConceptScore W4381586568C41008148 @default.
- W4381586568 hasConceptScore W4381586568C41895202 @default.
- W4381586568 hasConceptScore W4381586568C7149132 @default.
- W4381586568 hasConceptScore W4381586568C81917197 @default.
- W4381586568 hasConceptScore W4381586568C93959086 @default.
- W4381586568 hasConceptScore W4381586568C95623464 @default.
- W4381586568 hasLocation W43815865681 @default.
- W4381586568 hasOpenAccess W4381586568 @default.
- W4381586568 hasPrimaryLocation W43815865681 @default.
- W4381586568 hasRelatedWork W3158264953 @default.
- W4381586568 hasRelatedWork W3212218375 @default.
- W4381586568 hasRelatedWork W4223943233 @default.
- W4381586568 hasRelatedWork W4226365389 @default.
- W4381586568 hasRelatedWork W4310989423 @default.
- W4381586568 hasRelatedWork W4312200629 @default.
- W4381586568 hasRelatedWork W4312347790 @default.
- W4381586568 hasRelatedWork W4321153836 @default.
- W4381586568 hasRelatedWork W4360585206 @default.
- W4381586568 hasRelatedWork W4364306694 @default.
- W4381586568 isParatext "false" @default.
- W4381586568 isRetracted "false" @default.
- W4381586568 workType "article" @default.