Bidirectional Loss Function for Label Enhancement and Distribution Learning

@article{Liu2021BidirectionalLF,
  title={Bidirectional Loss Function for Label Enhancement and Distribution Learning},
  author={Xinyuan Liu and Jihua Zhu and Qinghai Zheng and Zhongyu Li and Ruixin Liu and Jun Wang},
  journal={Knowl. Based Syst.},
  year={2021},
  volume={213},
  pages={106690}
}
2 Citations

References

SHOWING 1-10 OF 46 REFERENCES

Label Distribution Learning by Exploiting Label Correlations

A novel label distribution learning algorithm to exploit correlations between different labels that performs remarkably better than both the state-of-the-art LDL methods and multi-label learning methods.

Latent Semantics Encoding for Label Distribution Learning

A novel algorithm is proposed, i.e., Latent Semantics Encoding forLabelDistributionLearning (LSE-LDL), which learns the label distribution and implements feature selection simultaneously under the guidance of latent semantics.

Label Enhancement for Label Distribution Learning

This paper proposes a novel LE algorithm called Graph Laplacian Label Enhancement (GLLE), which shows clear advantages over several existing LE algorithms and experimental results on eleven multi-label learning datasets validate the advantage of GLLE over the state-of-the-art multi- label learning approaches.

Label Distribution Learning

  • Xin Geng
  • Computer Science
    IEEE Transactions on Knowledge and Data Engineering
  • 2016
This paper proposes six working LDL algorithms in three ways: problem transformation, algorithm adaptation, and specialized algorithm design, and results show clear advantages of the specialized algorithms, which indicates the importance of special design for the characteristics of the LDL problem.

Label distribution learning with label-specific features

This paper proposes a novel LDL algorithm by leveraging label-specific features, where the common features for all labels and specific features for each label are simultaneously learned to enhance the LDL model.

Incomplete Label Distribution Learning

This paper proposes an objective based on trace norm minimization to exploit the correlation between labels and develops a proximal gradient descend algorithm and an algorithm based on alternating direction method of multipliers to solve LDL problem when given incomplete supervised information.

Label Distribution Learning with Label Correlations via Low-Rank Approximation

Both the global and local relevance among labels are utilized to provide more information for training model and a novel label distribution learning algorithm is proposed that outperforms state-of-the-art LDL methods.

Logistic Boosting Regression for Label Distribution Learning

  • Chao XingXin GengH. Xue
  • Computer Science
    2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2016
Experiments on facial expression recognition, crowd opinion prediction on movies and apparent age estimation show that LDLogitBoost and AOSO-LD logistic Boosting Regression can achieve better performance than traditional LDL algorithms as well as other LogitBoost algorithms.