Robust meta gradient learning for high-dimensional data with noisy-label ignorance

Robust meta gradient learning for high-dimensional data with noisy-label ignorance. Liu, B. & Lin, Y. PLoS ONE, 2023. Publisher: Public Library of Science Type: Article

Paper doi abstract bibtex

Large datasets with noisy labels and high dimensions have become increasingly prevalent in industry. These datasets often contain errors or inconsistencies in the assigned labels and introduce a vast number of predictive variables. Such issues frequently arise in real-world scenarios due to uncertainties or human errors during data collection and annotation processes. The presence of noisy labels and high dimensions can significantly impair the generalization ability and accuracy of trained models. To address the above issues, we introduce a simple-structured penalized γ-divergence model and a novel meta-gradient correction algorithm and establish the foundations of these two modules based on rigorous theoretical proofs. Finally, comprehensive experiments are conducted to validate their effectiveness in detecting noisy labels and mitigating the curse of dimensionality and suggest that our proposed model and algorithm can achieve promising outcomes. Moreover, we open-source our codes and distinctive datasets on GitHub (refer to https://github.com/DebtVC2022/Robust_Learning_with_MGC). © 2023 Public Library of Science. All rights reserved.

@article{liu_robust_2023,
	title = {Robust meta gradient learning for high-dimensional data with noisy-label ignorance},
	volume = {18},
	issn = {19326203},
	url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85179640429&doi=10.1371%2fjournal.pone.0295678&partnerID=40&md5=118b8b72caa109974bcf330e63f6d667},
	doi = {10.1371/journal.pone.0295678},
	abstract = {Large datasets with noisy labels and high dimensions have become increasingly prevalent in industry. These datasets often contain errors or inconsistencies in the assigned labels and introduce a vast number of predictive variables. Such issues frequently arise in real-world scenarios due to uncertainties or human errors during data collection and annotation processes. The presence of noisy labels and high dimensions can significantly impair the generalization ability and accuracy of trained models. To address the above issues, we introduce a simple-structured penalized γ-divergence model and a novel meta-gradient correction algorithm and establish the foundations of these two modules based on rigorous theoretical proofs. Finally, comprehensive experiments are conducted to validate their effectiveness in detecting noisy labels and mitigating the curse of dimensionality and suggest that our proposed model and algorithm can achieve promising outcomes. Moreover, we open-source our codes and distinctive datasets on GitHub (refer to https://github.com/DebtVC2022/Robust\_Learning\_with\_MGC). © 2023 Public Library of Science. All rights reserved.},
	language = {English},
	number = {12 December},
	journal = {PLoS ONE},
	author = {Liu, Ben and Lin, Yu},
	year = {2023},
	pmid = {38079441},
	note = {Publisher: Public Library of Science
Type: Article},
	keywords = {Algorithms, Data Collection, Generalization, Humans, Industry, Learning, Psychological, aged, algorithm, article, controlled study, drug therapy, generalization (psychology), human, industry, information processing, learning},
}

Downloads: 0