Graph classification based on optimizing graph spectra

Authors:
Nguyen Duy Vinh;Akihiro Inokuchi;Takashi Washio
Affiliations:
The Institute of Scientific and Industrial Research, Osaka University;The Institute of Scientific and Industrial Research, Osaka University and PRESTO, Japan Science and Technology Agency;The Institute of Scientific and Industrial Research, Osaka University
Venue:
DS'10 Proceedings of the 13th international conference on Discovery science
Year:
2010

Citing 7
Cited 0

The monotonicity theorem, Cauchy's interlace theorem, and the Courant-Fischer theorem

American Mathematical Monthly
A decision-theoretic generalization of on-line learning and an application to boosting

Journal of Computer and System Sciences - Special issue: 26th annual ACM symposium on the theory of computing & STOC'94, May 23–25, 1994, and second annual Europe an conference on computational learning theory (EuroCOLT'95), March 13–15, 1995
Computers and Intractability: A Guide to the Theory of NP-Completeness

Computers and Intractability: A Guide to the Theory of NP-Completeness
Kernel Methods for Pattern Analysis

Kernel Methods for Pattern Analysis
Kernels and Distances for Structured Data

Machine Learning
More generality in efficient multiple kernel learning

ICML '09 Proceedings of the 26th Annual International Conference on Machine Learning
A Linear-Time Graph Kernel

ICDM '09 Proceedings of the 2009 Ninth IEEE International Conference on Data Mining

Quantified Score

Hi-index	0.00

Visualization

Abstract

Kernel methods such as the SVM are becoming increasingly popular due to their high performance in graph classification. In this paper, we propose a novel graph kernel, called SPEC, based on graph spectra and the Interlace Theorem, as well as an algorithm, called OPTSPEC, to optimize the SPEC kernel used in an SVM for graph classification. The fundamental performance of the method is evaluated using artificial datasets, and its practicality confirmed through experiments using a real-world dataset.