Inferring network structure from cascades

Ghonge, Sushrut; Vural, Dervis Can

doi:10.1103/physreve.96.012319

Cited by 5 publications

(3 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many attempts have been made to solve this problem only based on diffusion knowledge, from static assumption requiring time stamp [15], [16], [17], [18] or without infection time [19] to dynamic inference [20], [21]. As a step further towards using prior knowledge, some works employ in-degree distribution for nodes [22], and measurements such as pathways, network properties, and information about the links or nodes [23]. Diffusion prediction: How to identify the future infected nodes in a cascade sequence by observing the incomplete and primary part of the cascade?…”

Section: Related Workmentioning

confidence: 99%

Joint Inference of Diffusion and Structure in Partially Observed Social Networks Using Coupled Matrix Factorization

Ramezani¹,

Ahadinia²,

Ziaei³

et al. 2020

Preprint

View full text Add to dashboard Cite

Access to complete data in large scale networks is often infeasible. Therefore, the problem of missing data is a crucial and unavoidable issue in analysis and modeling of real-world social networks. However, most of the research on different aspects of social networks do not consider this limitation. One effective way to solve this problem is to recover the missing data as a pre-processing step. The present paper tries to infer the unobserved data from both diffusion network and network structure by learning a model from the partially observed data. We develop a probabilistic generative model called "DiffStru" to jointly discover the hidden links of network structure and the omitted diffusion activities. The interrelations among links of nodes and cascade processes are utilized in the proposed method via learning coupled low dimensional latent factors. In addition to inferring the unseen data, the learned latent factors may also help network classification problems such as community detection. Simulation results on synthetic and real-world datasets show the excellent performance of the proposed method in terms of link prediction and discovering the identity and infection time of invisible social behaviors.

show abstract

Section: Related Workmentioning

confidence: 99%

Joint Inference of Diffusion and Structure in Partially Observed Social Networks Using Coupled Matrix Factorization

Ramezani¹,

Ahadinia²,

Ziaei³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Moreover, this might actually not be necessary if the goal is to estimate the global outcome of a contagion process at the population level and not the risk concerning a specific individual. Other approaches have been developed to specifically estimate important properties of epidemic spread and information cascade without trying to recover the original network [15,20,[27][28][29]. These methods are however either process specific or rely on the existence of known mesocale structures (such as groups in the population) in the network, together with the knowledge of the structure to which each population member belongs.…”

Section: Introductionmentioning

confidence: 99%

Estimating the outcome of spreading processes on networks with incomplete information: A dimensionality reduction approach

et al. 2018

View full text Add to dashboard Cite

Recent advances in data collection have facilitated the access to time-resolved human proximity data that can conveniently be represented as temporal networks of contacts between individuals. While the structural and dynamical information revealed by this type of data is fundamental to investigate how information or diseases propagate in a population, data often suffer from incompleteness, which possibly leads to biased estimations in data-driven models. A major challenge is thus to estimate the outcome of spreading processes occurring on temporal networks built from partial information. To cope with this problem, we devise an approach based on non-negative tensor factorization, a dimensionality reduction technique from multilinear algebra. The key idea is to learn a low-dimensional representation of the temporal network built from partial information and to use it to construct a surrogate network similar to the complete original network. To test our method, we consider several human-proximity networks, on which we perform resampling experiments to simulate a loss of data. Using our approach on the resulting partial networks, we build a surrogate version of the complete network for each. We then compare the outcome of a spreading process on the complete networks (nonaltered by a loss of data) and on the surrogate networks. We observe that the epidemic sizes obtained using the surrogate networks are in good agreement with those measured on the complete networks. Finally, we propose an extension of our framework that can leverage additional data, when available, to improve the surrogate network when the data loss is particularly large.

show abstract

“…Observable patterns here and now inform us of inaccessible patterns out and away. We can reconstruct the history of life from available fossils [4][5][6], or predict the fate of the universe by observing the present night sky [7][8][9]; we can infer hidden states and transition probabilities [10][11][12], connections and weights of neural networks [13][14][15] or parameters, initial states and interaction structures of complex systems [16][17][18][19][20][21][22][23][24][25].…”

mentioning

confidence: 99%

Enhancing the predictability and retrodictability of stochastic processes

Rupprecht

Vural

2019

Commun Phys

Self Cite

View full text Add to dashboard Cite

Scientific inference involves obtaining the unknown properties or behavior of a system in the light of what is known, typically, without changing the system. Here we propose an alternative to this approach: a system can be modified in a targeted way, preferably by a small amount, so that its properties and behavior can be inferred more successfully. For the sake of concreteness we focus on inferring the future and past of Markov processes and illustrate our method on two classes of processes: diffusion on random spatial networks, and thermalizing quantum systems. arXiv:1810.06620v3 [cond-mat.stat-mech]

show abstract

Inferring network structure from cascades

Cited by 5 publications

References 26 publications

Joint Inference of Diffusion and Structure in Partially Observed Social Networks Using Coupled Matrix Factorization

Joint Inference of Diffusion and Structure in Partially Observed Social Networks Using Coupled Matrix Factorization

Estimating the outcome of spreading processes on networks with incomplete information: A dimensionality reduction approach

Enhancing the predictability and retrodictability of stochastic processes

Contact Info

Product

Resources

About