BriefGPT.xyz
Oct, 2016
通过哈希加速连续属性图的核函数
Faster Kernels for Graphs with Continuous Attributes via Hashing
HTML
PDF
Christopher Morris, Nils M. Kriege, Kristian Kersting, Petra Mutzel
TL;DR
本文介绍了一种名为hash graph kernels的方法，可以将离散标签的图形核扩展到包含连续属性的图。该方法通过随机的哈希函数将连续属性迭代地转换成离散标签，并在Weisfeiler-Lehman子树核和最短路径核的基础上展示了其优越性能和扩展性。
Abstract
While state-of-the-art kernels for
graphs
with discrete labels scale well to
graphs
with thousands of nodes, the few existing kernels for
graphs<
→