The influence maximization problem seeks to maximize the spread of influence in network structures. This challenging optimization task remains unsolved due to its combinatorial nature and computational complexity. In this thesis, we propose two novel direct search methods, Neighbors Search (NS) and nonmonotone NS, which leverage the network structure to enhance efficiency. We compare these methods with the state-of-the-art Custom Direct Search (CDS) method through experiments on artificial and real-world networks. Our findings show promising improvements in solving the influence maximization problem.

The influence maximization problem seeks to maximize the spread of influence in network structures. This challenging optimization task remains unsolved due to its combinatorial nature and computational complexity. In this thesis, we propose two novel direct search methods, Neighbors Search (NS) and nonmonotone NS, which leverage the network structure to enhance efficiency. We compare these methods with the state-of-the-art Custom Direct Search (CDS) method through experiments on artificial and real-world networks. Our findings show promising improvements in solving the influence maximization problem.

Direct search methods for influence maximization problems

BERGAMASCHI, MATTEO
2022/2023

Abstract

The influence maximization problem seeks to maximize the spread of influence in network structures. This challenging optimization task remains unsolved due to its combinatorial nature and computational complexity. In this thesis, we propose two novel direct search methods, Neighbors Search (NS) and nonmonotone NS, which leverage the network structure to enhance efficiency. We compare these methods with the state-of-the-art Custom Direct Search (CDS) method through experiments on artificial and real-world networks. Our findings show promising improvements in solving the influence maximization problem.
2022
Direct search methods for influence maximization problems
The influence maximization problem seeks to maximize the spread of influence in network structures. This challenging optimization task remains unsolved due to its combinatorial nature and computational complexity. In this thesis, we propose two novel direct search methods, Neighbors Search (NS) and nonmonotone NS, which leverage the network structure to enhance efficiency. We compare these methods with the state-of-the-art Custom Direct Search (CDS) method through experiments on artificial and real-world networks. Our findings show promising improvements in solving the influence maximization problem.
Optimization
Graphs
Direct search method
File in questo prodotto:
File Dimensione Formato  
Matteo_Bergamaschi_Master_Thesis.pdf

accesso aperto

Dimensione 1.07 MB
Formato Adobe PDF
1.07 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/20.500.12608/52260