This thesis explores the local topology of the Geometric Random Intersection Graph (GRIG), a hybrid network model introduced by Deijfen and Michielan (2024) that blends spatial geometry with attribute-driven connectivity. The primary focus of this work is a rigorous analysis of the graph's clustering coefficient. Using Palm calculus and the properties of the underlying Poisson point processes, we derive an explicit analytical decomposition of the clustering coefficient into two distinct structural mechanisms: combinatorial clique-triangles, which form via shared group memberships, and geometric cycle-triangles, which arise from pure spatial proximity. We trace the behavior of this decomposition across different asymptotic parameter regimes, demonstrating how the clustering coefficient smoothly shifts between an attribute-sparse setting and an attribute-rich continuum percolation model. Finally, we validate these theoretical clustering limits through vectorized Monte Carlo simulations on a flat torus across several prototypical connection kernels.

The Geometric Random Intersection Graph Model: Clustering and Local Structure

DE NICOLÒ, ALBERTO
2025/2026

Abstract

This thesis explores the local topology of the Geometric Random Intersection Graph (GRIG), a hybrid network model introduced by Deijfen and Michielan (2024) that blends spatial geometry with attribute-driven connectivity. The primary focus of this work is a rigorous analysis of the graph's clustering coefficient. Using Palm calculus and the properties of the underlying Poisson point processes, we derive an explicit analytical decomposition of the clustering coefficient into two distinct structural mechanisms: combinatorial clique-triangles, which form via shared group memberships, and geometric cycle-triangles, which arise from pure spatial proximity. We trace the behavior of this decomposition across different asymptotic parameter regimes, demonstrating how the clustering coefficient smoothly shifts between an attribute-sparse setting and an attribute-rich continuum percolation model. Finally, we validate these theoretical clustering limits through vectorized Monte Carlo simulations on a flat torus across several prototypical connection kernels.
2025
The Geometric Random Intersection Graph Model: Clustering and Local Structure
GRIG
Clustering
Random graphs
Spatial graphs
Complex networks
File in questo prodotto:
File Dimensione Formato  
DeNicolò_Alberto.pdf

accesso aperto

Dimensione 3.33 MB
Formato Adobe PDF
3.33 MB Adobe PDF Visualizza/Apri

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/115675