Inferring Centrality from Network Snapshots

Shao, Haibin; Mesbahi, Mehran; Li, Dewei; Yang, Xi

doi:10.1038/srep40642

Cited by 13 publications

(7 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lastly, in terms of eigenvector centrality estimation from partial data, [25] considers the case of missing edges in the graph of interest, but does not rely on node data as we propose here. In the direction of centrality inference from data, [26] is the most similar to our work. However, the authors are concerned with temporal data driven by consensus dynamics, which they use to infer an ad hoc temporal centrality.…”

Section: Introductionsupporting

confidence: 52%

Blind Inference of Centrality Rankings from Graph Signals

Roddenberry

Segarra

2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

We study the blind centrality ranking problem, where our goal is to infer the eigenvector centrality ranking of nodes solely from nodal observations, i.e., without information about the topology of the network. We formalize these nodal observations as graph signals and model them as the outputs of a network process on the underlying (unobserved) network. A simple spectral algorithm is proposed to estimate the leading eigenvector of the associated adjacency matrix, thus serving as a proxy for the centrality ranking. A finite rate performance analysis of the algorithm is provided, where we find a lower bound on the number of graph signals needed to correctly rank (with high probability) two nodes of interest. We then specialize our general analysis for the particular case of dense Erdős-Rényi graphs, where existing graph-theoretical results can be leveraged. Finally, we illustrate the proposed algorithm via numerical experiments in synthetic and real-world networks, making special emphasis on how the network features influence the performance.

show abstract

Section: Introductionsupporting

confidence: 52%

Blind Inference of Centrality Rankings from Graph Signals

Roddenberry

Segarra

2020

ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…A wide range of multifaceted systems including social, biological, and technological networks display hierarchical organizations and exhibit universal topological features 6 , 38 , 39 . In these diverse systems, different centrality measures 40 including degree (number of links of a node), shortest paths (shortest distance between two nodes), and betweenness (number of shortest paths pass through a node), etc., might determine the transmission of information between two nodes in a connected network 41 . Given that effector targets exhibit an increased degree compared to non targets 31 – 33 , we hypothesized that diverse pathogens interact with effector targets to interfere with the flow of information.…”

Section: Resultsmentioning

confidence: 99%

Global temporal dynamic landscape of pathogen-mediated subversion of Arabidopsis innate immunity

Mishra

Sun

Ahmed

et al. 2017

Sci Rep

View full text Add to dashboard Cite

The universal nature of networks’ structural and physical properties across diverse systems offers a better prospect to elucidate the interplay between a system and its environment. In the last decade, several large-scale transcriptome and interactome studies were conducted to understand the complex and dynamic nature of interactions between Arabidopsis and its bacterial pathogen, Pseudomonas syringae pv. tomato DC3000. We took advantage of these publicly available datasets and performed “-omics”-based integrative, and network topology analyses to decipher the transcriptional and protein-protein interaction activities of effector targets. We demonstrated that effector targets exhibit shorter distance to differentially expressed genes (DEGs) and possess increased information centrality. Intriguingly, effector targets are differentially expressed in a sequential manner and make for 1% of the total DEGs at any time point of infection with virulent or defense-inducing DC3000 strains. We revealed that DC3000 significantly alters the expression levels of 71% effector targets and their downstream physical interacting proteins in Arabidopsis interactome. Our integrative “-omics”-–based analyses identified dynamic complexes associated with MTI and disease susceptibility. Finally, we discovered five novel plant defense players using a systems biology-fueled top-to-bottom approach and demonstrated immune-related functions for them, further validating the power and resolution of our network analyses.

show abstract

“…The relative tempo in Definition 2 was initially examined in [46], characterizing relative influence of agents in consensus networks, and subsequently being employed to construct a centrality measure that can be inferred from network data [47]. This paper provides a more systematic treatment for the application of relative tempo in the distributed neighbor selection problem.…”

Section: Neighbor Selection In Sansmentioning

confidence: 99%

Distributed Neighbor Selection in Multi-agent Networks

Shao

Pan

Mesbahi

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Achieving consensus via nearest neighbor rules is an important prerequisite for multi-agent networks to accomplish collective tasks. A common assumption in consensus setup is that each agent interacts with all its neighbors during the process. This paper examines whether network functionality and performance can be maintained-and even enhanced-when agents interact only with a subset of their respective (available) neighbors. As shown in the paper, the answer to this inquiry is affirmative. In this direction, we show that by using the monotonicity property of the Laplacian eigenvectors, a neighbor selection rule with guaranteed performance enhancements, can be realized for consensus-type networks. For the purpose of distributed implementation, a quantitative connection between Laplacian eigenvectors and the "relative rate of change" in the state between neighboring agents is further established; this connection facilitates a distributed algorithm for each agent to identify "favorable" neighbors to interact with. Multi-agent networks with and without external influence are examined, as well as extensions to signed networks. This paper underscores the utility of Laplacian eigenvectors in the context of distributed neighbor selection, providing novel insights into distributed datadriven control of multi-agent systems.

show abstract

Inferring Centrality from Network Snapshots

Cited by 13 publications

References 39 publications

Blind Inference of Centrality Rankings from Graph Signals

Blind Inference of Centrality Rankings from Graph Signals

Global temporal dynamic landscape of pathogen-mediated subversion of Arabidopsis innate immunity

Distributed Neighbor Selection in Multi-agent Networks

Contact Info

Product

Resources

About