TL;DR本文研究了深度图卷积网络由于 TRANS 操作引起的性能下降问题,通过提出一种节点归一化技术(NodeNorm)有效地缓解了这一问题。
Abstract
Graph Neural Networks (GNNs) tend to suffer performance degradation as model depth increases, which is usually attributed in previous works to the oversmoothing problem. However, we find that although oversmoothing is a contributing factor, the main reasons for this phenomenon are training difficulty and overfitting, which we study by experimentally investig