Proceedings of the 26th Annual International Conference on Machine Learning - ICML '09 2009
DOI: 10.1145/1553374.1553443
|View full text |Cite
|
Sign up to set email alerts
|

The graphlet spectrum

Abstract: Current graph kernels suffer from two limitations: graph kernels based on counting particular types of subgraphs ignore the relative position of these subgraphs to each other, while graph kernels based on algebraic methods are limited to graphs without node labels. In this paper we present the graphlet spectrum, a system of graph invariants derived by means of group representation theory that capture information about the number as well as the position of labeled subgraphs in a given graph. In our experimental… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
2
1

Citation Types

0
57
0

Year Published

2010
2010
2024
2024

Publication Types

Select...
5
3
2

Relationship

1
9

Authors

Journals

citations
Cited by 60 publications
(57 citation statements)
references
References 16 publications
0
57
0
Order By: Relevance
“…Typical applications include automated reasoning [36], bioinformatics/chemoinformatics [37, 38]. Generally speaking, graph kernels can be categorized into three classes: kernels based on walks [10, 11, 39, 40, 41], kernels based on limited-sized subgraphs [42, 43, 44] and kernels based on subtree patterns [45, 46, 47]. Graph kernels based on random walk is one of the most successful choices [48].…”
Section: Related Workmentioning
confidence: 99%
“…Typical applications include automated reasoning [36], bioinformatics/chemoinformatics [37, 38]. Generally speaking, graph kernels can be categorized into three classes: kernels based on walks [10, 11, 39, 40, 41], kernels based on limited-sized subgraphs [42, 43, 44] and kernels based on subtree patterns [45, 46, 47]. Graph kernels based on random walk is one of the most successful choices [48].…”
Section: Related Workmentioning
confidence: 99%
“…Kernels for discrete objects have been studied by Haussler [14], and many graph kernels have been proposed after that. There are three major groups in graph kernels: kernels based on walks and paths [9,20,10,3,42,43], kernels based on limited-size subgraphs [16,36,21], and kernels based on subtree patterns [25,35,15]. Among them, random walk based kernels have been proved to be one of the most successful methods [5,4,34].…”
Section: Related Workmentioning
confidence: 99%
“…We notice a few recent studies that are moving towards the direction of defining the relationship among features in graphs and sets. For example in the recently defined graph Graphlet Spectrum kernel [19], the spatial relationship of graph feature (called graphlets) are explored in an algebraic framework for measuring the structure similarity of graph adjacency matrices. In addition, recently developed association net uses a graph model to represent a set of association rules [29].…”
Section: Feature Graph Constructionmentioning
confidence: 99%