Noise-robust algorithm for graph data labeling
In brief
- GCNs performance drops with noisy node labels in real-world scenarios.
- LENS screens noisy nodes by measuring neighborhood entropy reduction.
- Boosts accuracy and speeds up training by tenfold compared to existing methods.
Graph Convolutional Networks rely on clean labels, a condition rarely met in real-world data. The proposed LENS method detects noisy nodes by analyzing neighborhood label distributions, enhancing classification accuracy while cutting computational costs by an order of magnitude.
This summary gives you the essentials. The full article, its figures and quotes are on cs.AI updates on arXiv.org.
Read the original articleAI-generated summary of the source, checked by rules. Report an error