2002 KernelsForGraphClassification

From GM-RKB
Jump to navigation Jump to search

Subject Headings: Kernel-based Learning Algorithm, Supervised Machine Learning Algorithm

Notes

Cited By

Quotes

Abstract

In this paper, we apply kernel methods to graph classification problems. To achieve the goal, we have to design an appropriate kernel for computing inner products for pairs of graphs represented in a feature space. We define a graph kernel by a random walk on a vertex product graph of two graphs. Some experiments on predicting properties of chemical compounds show encouraging results."

References


,

 AuthorvolumeDate ValuetitletypejournaltitleUrldoinoteyear
2002 KernelsForGraphClassificationHisashi Kashima
Akihiro Inokuchii
Kernels for graph classificationhttp://www.geocities.co.jp/Technopolis/5893/publication/am02.pdf