New MDTE framework boosts minority-class node classification in temporal graphs

Researchers have developed MDTE, a novel framework designed to improve node classification in temporal graphs, particularly for minority classes. This approach reconstructs temporal edge-event representations using conditional diffusion denoising, addressing the challenge of majority-dominated propagation that can obscure minority class signals. MDTE incorporates Distribution-Aware Selective Propagation to filter harmful propagation and Multi-View Discriminative Fusion to enhance class-wise distinctions. Experiments show MDTE significantly boosts minority-class recall and F1 scores compared to existing methods. AI

IMPACT Enhances the ability to identify and classify underrepresented data points in dynamic network structures.

RANK_REASON The cluster contains an academic paper detailing a new methodology for node classification. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv

• paper
• other

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