EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Influence Maximization in Complex Networks Using Deep Learning and Reinforcement Learning

Shiva Ahmadi Hormoghan, Bagher Zarei, Vahid Majidnezhad, Babak Anari

Complexity · 2026-01 · 2026 (1)

DOI: 10.1155/cplx/4861287 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Complexity, online erschienen am 26.06.2026. Verfasst von Shiva Ahmadi Hormoghan und 3 weiteren Autor:innen. Die Arbeit nennt 18 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Complex networks are made up of many components or nodes that interact with each other. These networks have recently been widely used and are popular platforms for product promotion and information diffusion. The interests and behavior of individuals in a complex network are highly influenced by each other. Finding individuals in complex networks as seed nodes of the network propagation is one of their most important problems, which exert the most significant influence on the remaining nodes in the network. Individuals are influential publishers who, if selected as the seed set in a publication issue in the network, will have the most people who have information about that published entity. This issue is referred to as the influence maximization (IM) problem. In various areas, such as recommender systems, IM has a significant commercial value in complex networks. In this research, we introduce an approach that integrates a graph convolutional network (GCN) with proximal policy optimization (PPO) to select a subset of the most influential nodes in a network with the objective of maximizing influence spread. This integrated approach enhances the ability to accurately detect nodes with the highest potential impact. The findings indicate that the proposed approach not only outperforms alternatives but also provides enhanced computational efficiency. The model’s capabilities are evaluated in comparison with existing approaches. Empirical evaluations on real‐world datasets further confirm that the proposed method achieves superior influence spread and enhanced performance in terms of runtime.

Zitieren

APA

Hormoghan, S. A., Zarei, B., Majidnezhad, V., & Anari, B. (2026). Influence Maximization in Complex Networks Using Deep Learning and Reinforcement Learning. Complexity, 2026(1). https://doi.org/10.1155/cplx/4861287

Vancouver

Hormoghan SA, Zarei B, Majidnezhad V, Anari B. Influence Maximization in Complex Networks Using Deep Learning and Reinforcement Learning. Complexity. 2026;2026(1). doi:10.1155/cplx/4861287

BibTeX
@article{Hormoghan2026,
  author = {Hormoghan, Shiva Ahmadi and Zarei, Bagher and Majidnezhad, Vahid and Anari, Babak},
  title = {Influence Maximization in Complex Networks Using Deep Learning and Reinforcement Learning},
  journal = {Complexity},
  year = {2026},
  volume = {2026},
  number = {1},
  doi = {10.1155/cplx/4861287}
}
Kostenlos starten →