Integration of Time Series and Multiplex Networks Series in Cybersecurity and Prediction: An Innovative Approach
Year: 2025
Type: book-chapter
Abstract: This review covers a recent and innovative methodological framework that uses novel tools developed within the field of network science to address some of the limitations of traditional time series and network analysis techniques. It summarizes and complements a set of studies that introduce and validate a novel approach based on the transformation of interaction-based time series into multiplex networks. This is achieved by first converting time series into visibility graphs (VGs), extracting their topological characteristics, grouping these characteristics to identify distinct behavior patterns, and finally structuring these patterns as layers within a multiplex network. At the core of this framework is the temporal metagraph, a mathematical structure in which temporal dynamics between entities are intrinsically encoded. The effectiveness of this framework is demonstrated by its application in two disparate and complex domains: cybersecurity, where it improves network intrusion detection systems (NIDSs) by providing more accurate and efficient identification of attackers; and predictive analytics, where it leverages urban mobility data to forecast real estate market trends with significant improvements over state-of-the-art models. The synthesis reveals that this integrated approach offers a powerful, generalizable, and computationally efficient paradigm for extracting useful intelligence from vast streams of temporal interaction data.
Source: IntechOpen eBooks
Authors: Sergio Iglesias Pérez, Regino Criado Herrero
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: hybrid
Awards: G2815892722, G546836605
OpenAlex ID: W4416208200