… both of which are large on multi-labeltext data, using only L2 … correlated features, possibly hurting the generalization — so L2 is still necessary for spreading weights across correlated …
R Song, Z Liu, X Chen, H An, Z Zhang, X Wang… - Applied Intelligence, 2023 - Springer
… correlations among labels in a complex and unknown label space. In this paper, we propose a Label Prompt Multi-labelTextClassification … -labeltextclassification, integrate labels into …
L Xu, Z Wang, Z Shen, Y Wang… - 2014 IEEE international …, 2014 - ieeexplore.ieee.org
… For example, in gene and protein function prediction, multiple functional labels are associated with each gene and protein [1]; in text categorization, a document can be assigned to …
… To make effective use of the labelcorrelations, we design a … methods for multi-labeltext classification, mainly including problem … of neural networks to multi-labelclassification problems. …
… labelcorrelation, we build a correlation matrix based on the cooccurrence of input labels. To make the matrix more efficient, we design a mechanism that normalizes the correlation …
A Garg, J Noyola, R Verma, A Saxena… - arXiv preprint arXiv …, 2015 - arxiv.org
… multi-labelclassification [1]. This paper extracts correlation information between labels and … options such as semantic relations in text to make classifications and context of user history. …
H Liu, C Yuan, X Wang - … Processing and Chinese Computing: 9th CCF …, 2020 - Springer
… To examine the ability of discovering the labelcorrelations, we cite three correlatedlabels “alternative”, “metal” and “rock” in music tagging task of RMSC-V2 dataset for example. In …
T Qian, F Li, M Zhang, G Jin, P Fan, W Dai - Neurocomputing, 2022 - Elsevier
… labelcorrelations, resulting in a significant improvement of the top-2 accuracy for textclassification (eg, 6.2… with fine-grained or a large number of labels. In particular, the LLCL is also an …
L Cai, Y Song, T Liu, K Zhang - Ieee Access, 2020 - ieeexplore.ieee.org
… labels, which caused information loss. In this paper, we show that explicitly modeling label semantics can improve multi-labeltextclassification. … the labelcorrelations by label graph built …