FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3196741022> ?p ?o ?g. }

• W3196741022 endingPage "63" @default.
• W3196741022 startingPage "55" @default.
• W3196741022 abstract "Graph neural networks (GNNs), which learn the node representations by recursively aggregating information from its neighbors, have become a predominant computational tool in many domains. To handle large-scale graphs, most of the existing methods partition the input graph into multiple sub-graphs (e.g., through node clustering) and apply batch training to save memory cost. However, such batch training will lead to label bias within each batch, and then result in over-confidence in model predictions. Since the connected nodes with positively related labels tend to be assigned together, the traditional cross-entropy minimization process will attend on the predictions of biased classes in the batch, and may intensify the overfitting issue. To overcome the label bias problem, we propose the adaptive label smoothing (ALS) method to replace the one-hot hard labels with smoothed ones, which learns to allocate label confidences from the biased classes to the others. Specifically, ALS propagates node labels to aggregate the neighborhood label distribution in a pre-processing step, and then updates the optimal smoothed labels online to adapt to specific graph structure. Experiments on the real-world datasets demonstrate that ALS can be generally applied to the main scalable learning frameworks to calibrate the biased labels and improve generalization performances." @default.
• W3196741022 created "2021-09-13" @default.
• W3196741022 creator A5007489034 @default.
• W3196741022 creator A5027475930 @default.
• W3196741022 creator A5027525032 @default.
• W3196741022 creator A5029252408 @default.
• W3196741022 creator A5053453125 @default.
• W3196741022 creator A5060076675 @default.
• W3196741022 creator A5068477431 @default.
• W3196741022 creator A5075644135 @default.
• W3196741022 date "2023-01-01" @default.
• W3196741022 modified "2023-09-27" @default.
• W3196741022 title "Adaptive Label Smoothing To Regularize Large-Scale Graph Training" @default.
• W3196741022 cites W1630959083 @default.
• W3196741022 cites W1662382123 @default.
• W3196741022 cites W2019512103 @default.
• W3196741022 cites W2097777089 @default.
• W3196741022 cites W2104290444 @default.
• W3196741022 cites W2136504847 @default.
• W3196741022 cites W2139823104 @default.
• W3196741022 cites W2141088152 @default.
• W3196741022 cites W2154455818 @default.
• W3196741022 cites W2169884321 @default.
• W3196741022 cites W2183341477 @default.
• W3196741022 cites W2256388387 @default.
• W3196741022 cites W2296073425 @default.
• W3196741022 cites W2296407087 @default.
• W3196741022 cites W2345474290 @default.
• W3196741022 cites W2468907370 @default.
• W3196741022 cites W2519887557 @default.
• W3196741022 cites W2606780347 @default.
• W3196741022 cites W2624431344 @default.
• W3196741022 cites W2766453196 @default.
• W3196741022 cites W2781957615 @default.
• W3196741022 cites W2786915849 @default.
• W3196741022 cites W2790814121 @default.
• W3196741022 cites W2884014654 @default.
• W3196741022 cites W2894175828 @default.
• W3196741022 cites W2914721378 @default.
• W3196741022 cites W2916106175 @default.
• W3196741022 cites W2948210185 @default.
• W3196741022 cites W2950300355 @default.
• W3196741022 cites W2951659295 @default.
• W3196741022 cites W2953406194 @default.
• W3196741022 cites W2961295589 @default.
• W3196741022 cites W2963601856 @default.
• W3196741022 cites W2964124573 @default.
• W3196741022 cites W2968685800 @default.
• W3196741022 cites W2971564650 @default.
• W3196741022 cites W2972672017 @default.
• W3196741022 cites W2994872218 @default.
• W3196741022 cites W2998431760 @default.
• W3196741022 cites W3005104128 @default.
• W3196741022 cites W3019011053 @default.
• W3196741022 cites W3021708827 @default.
• W3196741022 cites W3022579377 @default.
• W3196741022 cites W3034492151 @default.
• W3196741022 cites W3034618665 @default.
• W3196741022 cites W3034756453 @default.
• W3196741022 cites W3035659929 @default.
• W3196741022 cites W3036585823 @default.
• W3196741022 cites W3084428871 @default.
• W3196741022 cites W3095794553 @default.
• W3196741022 cites W3097264851 @default.
• W3196741022 cites W3100848837 @default.
• W3196741022 cites W3104097132 @default.
• W3196741022 cites W3123816584 @default.
• W3196741022 cites W3156681329 @default.
• W3196741022 cites W3156994078 @default.
• W3196741022 cites W3173623546 @default.
• W3196741022 cites W3178159063 @default.
• W3196741022 cites W3021975806 @default.
• W3196741022 cites W3037079993 @default.
• W3196741022 doi "https://doi.org/10.1137/1.9781611977653.ch7" @default.
• W3196741022 hasPublicationYear "2023" @default.
• W3196741022 type Work @default.
• W3196741022 sameAs 3196741022 @default.
• W3196741022 citedByCount "1" @default.
• W3196741022 countsByYear W31967410222021 @default.
• W3196741022 crossrefType "book-chapter" @default.
• W3196741022 hasAuthorship W3196741022A5007489034 @default.
• W3196741022 hasAuthorship W3196741022A5027475930 @default.
• W3196741022 hasAuthorship W3196741022A5027525032 @default.
• W3196741022 hasAuthorship W3196741022A5029252408 @default.
• W3196741022 hasAuthorship W3196741022A5053453125 @default.
• W3196741022 hasAuthorship W3196741022A5060076675 @default.
• W3196741022 hasAuthorship W3196741022A5068477431 @default.
• W3196741022 hasAuthorship W3196741022A5075644135 @default.
• W3196741022 hasBestOaLocation W31967410222 @default.
• W3196741022 hasConcept C119857082 @default.
• W3196741022 hasConcept C124101348 @default.
• W3196741022 hasConcept C132525143 @default.
• W3196741022 hasConcept C154945302 @default.
• W3196741022 hasConcept C22019652 @default.
• W3196741022 hasConcept C31972630 @default.
• W3196741022 hasConcept C3770464 @default.
• W3196741022 hasConcept C41008148 @default.
• W3196741022 hasConcept C48044578 @default.

• next