BriefGPT.xyz
Jun, 2019
Wasserstein Weisfeiler-Lehman图核
Wasserstein Weisfeiler-Lehman Graph Kernels
HTML
PDF
Matteo Togninalli, Elisabetta Ghisu, Felipe Llinares-López, Bastian Rieck, Karsten Borgwardt
TL;DR
本文介绍了一种基于Wasserstein距离和Weisfeiler-Lehman embedding的新型图卷积核方法,以图像高维对象的方式比较图结构之间的相似性并在多个图分类任务上提高了预测性能。
Abstract
graph kernels
are an instance of the class of $\mathcal{R}$-Convolution kernels, which measure the similarity of objects by comparing their substructures. Despite their empirical success, most
graph kernels
use a
→