By the same authors

An Aligned Subtree Kernel for Weighted Graphs

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Full text download(s)

  • bai15

    324 KB, PDF document

Author(s)

Department/unit(s)

Publication details

Title of host publicationInternational Conference on Machine Learning (ICML) 2015
DatePublished - 2015
Pages30-39
Number of pages10
Volume37
Original languageEnglish

Abstract


In this paper, we develop a new entropic match- ing kernel for weighted graphs by aligning depth- based representations. We demonstrate that this kernel can be seen as an aligned subtree kernel that incorporates explicit subtree correspondences, and thus addresses the drawback of neglecting the relative locations between substructures that arises in the R-convolution kernels. Experiments on standard datasets demonstrate that our kernel can easily outperform state-of-the-art graph kernels in terms of classification accuracy.

Discover related content

Find related publications, people, projects, datasets and more using interactive charts.

View graph of relations