PaSta framework improves noisy node classification using partial label learning

Researchers have introduced PaSta, a novel framework for noisy node classification that utilizes partial label learning. This method addresses the limitations of existing approaches by training multiple annotators to generate high-quality partial labels, which are then used to guide a classification model. PaSta further enhances robustness through a self-training strategy, iteratively refining labels and optimizing annotators. Experiments show PaSta achieves an average improvement of 1.1% in classification performance across various noise settings. AI

IMPACT This research offers a novel approach to improve the accuracy of graph-based machine learning models in real-world scenarios with imperfect data.

RANK_REASON The item is an academic paper detailing a new method for noisy node classification. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• IArxiv Recommender
• Influence Flower
• Partial Label Learning with competitive learning graph neural network
• PaSta
• ScienceCast

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →