Degree Set Characterization and Circulant Graph-Based Pruning for Neural Networks
Boletim da Sociedade Paranaense de Matemática · 30.07.2026 · 44 (8) · 1-10
DOI: 10.5269/bspm.83828 ↗Kurz gefasst
Fachartikel in der Zeitschrift Boletim da Sociedade Paranaense de Matemática, online erschienen am 30.07.2026. Verfasst von M. VijayaLakshmi und 1 weiteren Autor:innen. Die Arbeit nennt 29 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
The degree set D(G) of a graph G is the collection of all vertex degrees in the graph. For a given positive integer n, criteria for the existence of multipartite graphs whose degree set is {n} are established. Enumeration problems concerning multipartite graphs whose degree sets consist of primes or products of primes are solved. Additionally, a characterization of multipartite graphs whose degree sets consist of specified positive integers is provided and the minimum order required for such graphs is determined. As an extension, we apply the combinatorial frame work to neural network pruning by modeling layers as circulant graphs C_n(S). This structure enables degree-based pruning using weighted degrees to detect low-utility neurons. The method promotes sparsity and interpretability while preserving performance, offering a graph-theoretic approach to efficient model compression.
Zitieren
APA
VijayaLakshmi, M., & Raja, V. (2026). Degree Set Characterization and Circulant Graph-Based Pruning for Neural Networks. Boletim da Sociedade Paranaense de Matemática, 44(8), 1-10. https://doi.org/10.5269/bspm.83828
Vancouver
VijayaLakshmi M, Raja V. Degree Set Characterization and Circulant Graph-Based Pruning for Neural Networks. Boletim da Sociedade Paranaense de Matemática. 2026;44(8):1-10. doi:10.5269/bspm.83828
BibTeX
@article{VijayaLakshmi2026,
author = {VijayaLakshmi, M. and Raja, V.},
title = {Degree Set Characterization and Circulant Graph-Based Pruning for Neural Networks},
journal = {Boletim da Sociedade Paranaense de Matemática},
year = {2026},
volume = {44},
number = {8},
pages = {1--10},
doi = {10.5269/bspm.83828}
}
EUVIMED