Higher-order interactions shape collective dynamics differently in hypergraphs and simplicial complexes

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: March 23, 2023

Abstract Higher-order networks have emerged as a powerful framework to model complex systems and their collective behavior. Going beyond pairwise interactions, they encode structured relations among arbitrary numbers of units through representations such simplicial complexes hypergraphs. So far, the choice between hypergraphs has often been motivated by technical convenience. Here, using synchronization an example, we demonstrate that effects higher-order interactions are highly representation-dependent. In particular, typically enhance in but opposite effect complexes. We provide theoretical insight linking synchronizability different hypergraph structures (generalized) degree heterogeneity cross-order correlation, which turn influence wide range dynamical processes from contagion diffusion. Our findings reveal hidden impact on dynamics, highlighting importance choosing appropriate when studying with nonpairwise interactions.

Language: Английский

---

Partial entropy decomposition reveals higher-order information structures in human brain activity

Proceedings of the National Academy of Sciences, Journal Year: 2023, Volume and Issue: 120(30)

Published: July 19, 2023

The standard approach to modeling the human brain as a complex system is with network, where basic unit of interaction pairwise link between two regions. While powerful, this limited by inability assess higher-order interactions involving three or more elements directly. In work, we explore method for capturing dependencies in multivariate data: partial entropy decomposition (PED). Our decomposes joint whole into set nonnegative atoms that describe redundant, unique, and synergistic compose system's structure. PED gives insight mathematics functional connectivity its limitation. When applied resting-state fMRI data, find robust evidence synergies are largely invisible analyses. can also be localized time, allowing frame-by-frame analysis how distributions redundancies change over course recording. We different ensembles regions transiently from being redundancy-dominated synergy-dominated temporal pattern structured time. These results provide strong there exists large space unexplored structures data have been missed focus on bivariate network models. This structure dynamic time likely will illuminate interesting links behavior. Beyond brain-specific application, provides very general understanding variety systems.

Language: Английский

---

Community detection in large hypergraphs

Science Advances, Journal Year: 2023, Volume and Issue: 9(28)

Published: July 12, 2023

Hypergraphs, describing networks where interactions take place among any number of units, are a natural tool to model many real-world social and biological systems. Here, we propose principled framework the organization higher-order data. Our approach recovers community structure with accuracy exceeding that currently available state-of-the-art algorithms, as tested in synthetic benchmarks both hard overlapping ground-truth partitions. is flexible allows capturing assortative disassortative structures. Moreover, our method scales orders magnitude faster than competing making it suitable for analysis very large hypergraphs, containing millions nodes thousands nodes. work constitutes practical general hypergraph analysis, broadening understanding

Language: Английский

---

Contagion dynamics on higher-order networks

Nature Reviews Physics, Journal Year: 2024, Volume and Issue: 6(8), P. 468 - 482

Published: July 5, 2024

Language: Английский

---

Advancing urban traffic accident forecasting through sparse spatio-temporal dynamic learning

Accident Analysis & Prevention, Journal Year: 2024, Volume and Issue: 200, P. 107564 - 107564

Published: April 2, 2024

Language: Английский

---

Local dominance unveils clusters in networks

Communications Physics, Journal Year: 2024, Volume and Issue: 7(1)

Published: May 31, 2024

Abstract Clusters or communities can provide a coarse-grained description of complex systems at multiple scales, but their detection remains challenging in practice. Community methods often define as dense subgraphs, subgraphs with few connections in-between, via concepts such the cut, conductance, modularity. Here we consider another perspective built on notion local dominance, where low-degree nodes are assigned to basin influence high-degree nodes, and design an efficient algorithm based information. Local dominance gives rises community centers, uncovers hierarchies network. centers have larger degree than neighbors sufficiently distant from other centers. The strength our framework is demonstrated synthesized empirical networks ground-truth labels. associated asymmetric relations between not restricted detection, be utilised clustering problems, illustrate derived vector data.

Language: Английский

---

Higher-order correlations reveal complex memory in temporal hypergraphs

Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)

Published: June 4, 2024

Abstract Many real-world complex systems are characterized by interactions in groups that change time. Current temporal network approaches, however, unable to describe group dynamics, as they based on pairwise only. Here, we use time-varying hypergraphs such systems, and introduce a framework higher-order correlations characterize their organization. The analysis of human interaction data reveals the existence coherent interdependent mesoscopic structures, thus capturing aggregation, fragmentation nucleation processes social systems. We model with non-Markovian interactions, which memory fundamental mechanism underlying emerging pattern data.

Language: Английский

---

The time-evolving epileptic brain network: concepts, definitions, accomplishments, perspectives

Frontiers in Network Physiology, Journal Year: 2024, Volume and Issue: 3

Published: Jan. 16, 2024

Epilepsy is now considered a network disease that affects the brain across multiple levels of spatial and temporal scales. The paradigm shift from an epileptic focus—a discrete cortical area which seizures originate—to widespread network—spanning lobes hemispheres—considerably advanced our understanding epilepsy continues to influence both research clinical treatment this multi-faceted high-impact neurological disorder. network, however, not static but evolves in time requires novel approaches for in-depth characterization. In review, we discuss conceptual basics theory critically examine state-of-the-art recording techniques analysis tools used assess characterize time-evolving human network. We give account on current shortcomings highlight potential developments towards improved management epilepsy.

Language: Английский

---

Inference and visualization of community structure in attributed hypergraphs using mixed-membership stochastic block models

Social Network Analysis and Mining, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 10, 2025

Language: Английский

---

Hypergraphx: a library for higher-order network analysis

Journal of Complex Networks, Journal Year: 2023, Volume and Issue: 11(3)

Published: April 21, 2023

Abstract From social to biological systems, many real-world systems are characterized by higher-order, non-dyadic interactions. Such conveniently described hypergraphs, where hyperedges encode interactions among an arbitrary number of units. Here, we present open-source python library, hypergraphx (HGX), providing a comprehensive collection algorithms and functions for the analysis higher-order networks. These include different ways convert data across distinct representations, large variety measures organization at local mesoscale, statistical filters sparsify data, wide array static dynamic generative models, implementation dynamical processes with Our computational framework is general, allows analyse hypergraphs weighted, directed, signed, temporal multiplex group We provide visual insights on through visualization tools. accompany our code extended repository demonstrate ability HGX systematic network The library conceived as evolving, community-based effort, which will further extend its functionalities over years. software available https://github.com/HGX-Team/hypergraphx.

Language: Английский

---