Abstract
Purchase PDF (342 K)
E-mail Article
Add to my Quick Links
Cited By in Scopus (1)
Purchase PDF (1153 K)
doi:10.1016/j.neucom.2004.01.189 
Copyright © 2004 Elsevier B.V. All rights reserved.
Solving inexact graph isomorphism problems using neural networks
Received 23 October 2003;
accepted 23 January 2004.
Available online 19 August 2004.
References and further reading may be available for this article. To view references and further reading you must purchase this article.
Abstract
We present a neural network approach to solve exact and inexact graph isomorphism problems for weighted graphs. In contrast to other neural heuristics or related methods this approach is based on a neural refinement procedure to reduce the search space followed by an energy-minimizing matching process. Experiments on random weighted graphs in the range of 100–5000 vertices and on chemical molecular structures are presented and discussed.
Keywords: Graph isomorphism; Association graph; Maximum clique; Hopfield network
Article Outline
- 1. Introduction
- 2. Preliminaries
- 3. Exact and inexact vertex invariants
- 4. A neural
-GIP solver - 4.1. Step 1—Approximating the -automorphism partition
- 4.1.1. Exact case
- 4.1.2. Inexact case
- 4.1.3. Limitations
- 4.2. Step 2—Construction of an -association graph
- 4.3. Step 3—Solving the maximum clique problem using Hopfield networks
- 5. Experimental results
- 5.1. Random binary graphs
- 5.2. Chemical molecules
- 5.3. Random weighted graphs
- 6. Conclusion
- References
- Vitae
