Sparse Power-Law Network Model for Reliable Statistical Predictions Based on Sampled Data

@article{KartunGiles2018SparsePN,
  title={Sparse Power-Law Network Model for Reliable Statistical Predictions Based on Sampled Data},
  author={Alexander P. Kartun-Giles and Dmitri V. Krioukov and James P. Gleeson and Yamir Moreno and Ginestra Bianconi},
  journal={Entropy},
  year={2018},
  volume={20}
}
A projective network model is a model that enables predictions to be made based on a subsample of the network data, with the predictions remaining unchanged if a larger sample is taken into consideration. An exchangeable model is a model that does not depend on the order in which nodes are sampled. Despite a large variety of non-equilibrium (growing) and equilibrium (static) sparse complex network models that are widely used in network science, how to reconcile sparseness (constant average… 

Figures from this paper

Projective, Sparse, and Learnable Latent Position Network Models
TLDR
This work proposes an alternative assumption---that the latent positions are generated according to a Poisson point process---and shows that it is compatible with various levels of sparsity, thus ensuring consistency of statistical inference across networks of different sizes.
Introducing Graph Cumulants: What is the Kurtosis of your Social Network?
In an increasingly interconnected world, understanding and summarizing the structure of these networks becomes increasingly relevant. However, this task is nontrivial; proposed summary statistics are
Statistical physics of exchangeable sparse simple networks, multiplex networks, and simplicial complexes.
TLDR
A statistical physics framework and a Metropolis-Hastings algorithm defining exchangeable sparse network ensembles are proposed and both uncorrelated and correlated exchangeable ensembleles of simplicial complexes are treated using statistical mechanics approaches.
Connectivity of 1d random geometric graphs
TLDR
An important link between spatial random graphs, and lattice path combinatorics, where the d-dimensional lattice paths correspond to spatial permutations of the geometric points on the line is demonstrated and described.
Introducing Graph Cumulants: What is the Variance of Your Social Network?
1st order 2nd order 3rd order 1 Introducing Cumulants for Graphs: What is the Kurtosis of your Social Network? Gecia Bravo-Hermsdorff⇤,1 & Lee M. Gunderson⇤,2 ⇤Both authors contributed equally to
Development of the information system for HIV progression modeling in inmates
TLDR
It was estimated that the developed model gives the best results for patients under the age of 40 years and will help to predict correctly the HIV severity and to take timely treatment measures.

References

SHOWING 1-10 OF 85 REFERENCES
Edge exchangeable models for network data
TLDR
It is observed that edges, not vertices, act as the statistical units in networks constructed from interaction data, making a theory of edge-labeled networks more natural for many applications, and gives rise to a class of nonparametric models, akin to graphon models in the vertex exchangeable setting.
CONSISTENCY UNDER SAMPLING OF EXPONENTIAL RANDOM GRAPH MODELS.
TLDR
It is shown that this apparently trivial condition is in fact violated by many popular and scientifically appealing models, and that satisfying it drastically limits ERGM's expressive power.
Hierarchical block structures and high-resolution model selection in large networks
TLDR
A nested generative model is constructed that, through a complete description of the entire network hierarchy at multiple scales, enables the detection of modular structure at levels far beyond those possible with current approaches, and is based on the principle of parsimony.
Sparse graphs using exchangeable random measures
  • F. Caron, E. Fox
  • Computer Science
    Journal of the Royal Statistical Society. Series B, Statistical methodology
  • 2017
TLDR
A scalable Hamiltonian Monte Carlo algorithm for posterior inference is presented, which is used to analyse network properties in a range of real data sets, including networks with hundreds of thousands of nodes and millions of edges.
Maximum likelihood: extracting unbiased information from complex networks.
TLDR
This work shows that the maximum likelihood (ML) principle indicates a unique, statistically rigorous parameter choice, associated with a well-defined topological feature, and builds a class of safely unbiased models to overcome this problem.
New Specifications for Exponential Random Graph Models
TLDR
It is concluded that the new specifications of exponential random graph models increase the range and applicability of the ERGM as a tool for the statistical analysis of social networks.
The entropy of randomized network ensembles
TLDR
This paper characterize these randomized network ensembles by their entropy, i.e. the normalized logarithm of the total number of networks which are part of these ensemble, and proposes entropy as an indicator to assess the role of each structural feature in a given real network.
Statistical mechanics of networks.
  • Juyong Park, M. Newman
  • Computer Science
    Physical review. E, Statistical, nonlinear, and soft matter physics
  • 2004
We study the family of network models derived by requiring the expected properties of a graph ensemble to match a given set of measurements of a real-world network, while maximizing the entropy of
Sparse Maximum-Entropy Random Graphs with a Given Power-Law Degree Distribution
TLDR
It is proved that the hypersoft configuration model, belonging to the class of random graphs with latent hyperparameters, is an ensemble of random power-law graphs that are sparse, unbiased, and either exchangeable or projective.
Statistical mechanics of multiedge networks.
TLDR
A statistical mechanics framework is developed where it is possible to get information about the most relevant observables given a large spectrum of linear and nonlinear constraints including those depending both on the number of multiedges per link and their binary projection.
...
...