A Generalized Information-Theoretic Approach for Bounding the Number of Independent Sets in Bipartite Graphs.

Shannon entropy Shearer’s lemma counting graphs. independent sets

Journal

Entropy (Basel, Switzerland)
ISSN: 1099-4300
Titre abrégé: Entropy (Basel)
Pays: Switzerland
ID NLM: 101243874

Informations de publication

Date de publication:
25 Feb 2021
Historique:
received: 22 12 2020
revised: 18 02 2021
accepted: 22 02 2021
entrez: 6 3 2021
pubmed: 7 3 2021
medline: 7 3 2021
Statut: epublish

Résumé

This paper studies the problem of upper bounding the number of independent sets in a graph, expressed in terms of its degree distribution. For bipartite regular graphs, Kahn (2001) established a tight upper bound using an information-theoretic approach, and he also conjectured an upper bound for general graphs. His conjectured bound was recently proved by Sah et al. (2019), using different techniques not involving information theory. The main contribution of this work is the extension of Kahn's information-theoretic proof technique to handle irregular bipartite graphs. In particular, when the bipartite graph is regular on one side, but may be irregular on the other, the extended entropy-based proof technique yields the same bound as was conjectured by Kahn (2001) and proved by Sah et al. (2019).

Identifiants

pubmed: 33668754
pii: e23030270
doi: 10.3390/e23030270
pmc: PMC7996360
pii:
doi:

Types de publication

Journal Article

Langues

eng

Auteurs

Igal Sason (I)

Department of Electrical Engineering, Technion-Israel Institute of Technology, Haifa 3200003, Israel.

Classifications MeSH