arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2011.10988cs.LG
收藏

堆叠图过滤器

Stacked Graph Filter

Tokyo Tech(东京工业大学);RIKEN AIP(理化学研究所AIP)

Hoang NT, Takanori Maehara, Tsuyoshi Murata

摘要:

我们研究图卷积网络(GCN)从图形信号处理的观点,通过解决学习图滤波器与全连接的权重与可训练的多项式系数之间的差异。我们发现,通过堆叠图过滤器与可学习的多项式参数,我们可以建立一个高度自适应和强大的顶点分类模型。我们在这里的治疗放松了现有顶点分类模型中的低频(或等价地,高同质性)假设,从而在光谱特性方面得到了更普遍的解决方案。从经验上讲,通过仅使用一个超参数设置,我们的模型在频谱上的大多数基准数据集上都取得了很好的结果。

英文摘要:

We study Graph Convolutional Networks (GCN) from the graph signal processing viewpoint by addressing a difference between learning graph filters with fully connected weights versus trainable polynomial coefficients. We find that by stacking graph filters with learnable polynomial parameters, we can build a highly adaptive and robust vertex classification model. Our treatment here relaxes the low-frequency (or equivalently, high homophily) assumptions in existing vertex classification models, resulting a more ubiquitous solution in terms of spectral properties. Empirically, by using only one hyper-parameter setting, our model achieves strong results on most benchmark datasets across the frequency spectrum.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑