Please use this identifier to cite or link to this item:
http://hdl.handle.net/1893/18447
Appears in Collections: | Computing Science and Mathematics Research Reports |
Peer Review Status: | Refereed |
Title: | Spectral Reconstruction and Isomorphism of graphs using variable neighbourhood search |
Author(s): | Caporossi, Gilles Cvetkovic, Dragos Rowlinson, Peter |
Contact Email: | peter.rowlinson@stir.ac.uk |
Citation: | Caporossi G, Cvetkovic D & Rowlinson P (2013) Spectral Reconstruction and Isomorphism of graphs using variable neighbourhood search. Les Cahiers du GERAD, G-2013-73. GERAD. |
Keywords: | Spectral distance graph angles graph isomorphism variable neighbourhood search |
Issue Date: | Oct-2013 |
Publisher: | GERAD |
Series/Report no.: | Les Cahiers du GERAD, G-2013-73 |
Abstract: | The Euclidean distance between the eigenvalue sequences of graphs G and H, on the same number of vertices, is called the spectral distance between G and H. This notion is the basis of a heuristic algorithm for reconstructing a graph with prescribed spectrum. By using a graph Γ constructed from cospectral graphs G and H, we can ensure that G and H are isomorphic if and only if the spectral distance between Γ and G+K 2 is zero. This construction is exploited to design a heuristic algorithm for testing graph isomorphism. We present preliminary experimental results obtained by implementing these algorithms in conjunction with a meta-heuristic known as a variable neighbourhood search. |
Type: | Research Report |
URL: | http://www.gerad.ca/fichiers/cahiers/G-2013-73.pdf |
URI: | http://hdl.handle.net/1893/18447 |
Rights: | Authors retain copyright. |
Affiliation: | HEC Montreal Mathematical Institute SANU Mathematics - CSM Dept |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
G-2013-73.pdf | 336.23 kB | Adobe PDF | View/Open |
This item is protected by original copyright |
Items in the Repository are protected by copyright, with all rights reserved, unless otherwise indicated.
If you believe that any material held in STORRE infringes copyright, please contact library@stir.ac.uk providing details and we will remove the Work from public display in STORRE and investigate your claim.