.png){kind=icon}
.png){kind=icon}
Return
Frequent approximate subgraphs as features for graph-based image classification
.png){kind=icon}
PRE
AI
DOI：10.1016/j.knosys.2011.12.002.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The use of approximate graph matching for frequent subgraph mining has been identified in different applications as a need. To meet this need, several algorithms have been developed, but there are applications where it has not been used yet, for example image classification. In this paper, a new algorithm for mining frequent connected subgraphs over undirected and labeled graph collections VEAM (Vertex and Edge Approximate graph Miner) is presented. Slight variations of the data, keeping the topology of the graphs, are allowed in this algorithm. Approximate matching in existing algorithm (APGM) is only performed on vertex label set. In VEAM, the approximate matching between edge label set in frequent subgraph mining is included in the mining process. Also, a framework for graph-based image classification is introduced. The approximate method of VEAM was tested on an artificial image collection using a graph-based image representation proposed in this paper. The experimentation on this collection shows that our proposal gets better results than graph-based image classification using some algorithms reported in related work. (C) 2011 Elsevier B.V. All rights reserved.
Keywords:
Approximate graph mining
Approximate graph matching
Image representation
Image classification
Feature selection
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
K
IF:
7.6
Papers:
1.2W
Citations:
4.5W
Organization
No organization information available