Statistical validation of high-dimensional models of growing networks

@article{Medo2014StatisticalVO,
  title={Statistical validation of high-dimensional models of growing networks},
  author={Mat{\'u}{\vs} Medo},
  journal={Physical review. E, Statistical, nonlinear, and soft matter physics},
  year={2014},
  volume={89 3},
  pages={
          032801
        }
}
  • M. Medo
  • Published 8 November 2013
  • Computer Science
  • Physical review. E, Statistical, nonlinear, and soft matter physics
The abundance of models of complex networks and the current insufficient validation standards make it difficult to judge which models are strongly supported by data and which are not. We focus here on likelihood maximization methods for models of growing networks with many parameters and compare their performance on artificial and real datasets. While high dimensionality of the parameter space harms the performance of direct likelihood maximization on artificial data, this can be improved by… 

Figures and Tables from this paper

Ranking nodes in growing networks: When PageRank fails

TLDR
It is shown that realistic temporal effects make PageRank fail in individuating the most valuable nodes for a broad range of model parameters, and that time-dependent algorithms that are based on the temporal linking patterns of these systems are needed to better rank the nodes.

Optimal timescale for community detection in growing networks

TLDR
It is shown that before attempting to assess structural patterns of evolving networks, it is vital to uncover the timescales of the dynamical processes that generated them, and a multi-layer quality function is used to show this.

Ranking in evolving complex networks

A multi-layer network approach to modelling authorship influence on citation dynamics in physics journals

TLDR
A general method for statistical parameter estimation and model selection that is applicable to growing multilayer networks and takes both the parameter errors and the model complexity into account and is computationally efficient and scalable to large networks.

Structure-oriented prediction in complex networks

Mechanisms of complex network growth: Synthesis of the preferential attachment and fitness models.

TLDR
The fitness model yields the long-sought explanation for the initial attractivity K_{0}, an elusive parameter which was left unexplained within the framework of the preferential attachment model, and it is shown that theInitial attractivity is determined by the width of the fitness distribution.

Decay of Relevance in Exponentially Growing Networks

TLDR
A new preferential attachment-based network growth model is proposed in order to explain two properties of growing networks: (1) the power-law growth of node degrees and (2) the decay of node relevance, and it is found that apart from being empirically observed in many systems, the exponential growth of the network size over time is the key to sustain thePower-laws growth of nodes degrees when node relevance decays.

Phase Transition in the Recoverability of Network History

TLDR
A Bayesian formulation of network archaeology is introduced, with a generalization of preferential attachment as the generative mechanism, and a sequential importance sampling algorithm to evaluate the posterior averages of this model, as well as an efficient heuristic that uncovers the history of a network in linear time.

Time-invariant degree growth in preferential attachment network models

TLDR
A novel analytical framework is proposed based on the time-invariance of the studied systems and it is shown that it is self-consistent only for two special network growth forms: the uniform and exponential network growth.

References

SHOWING 1-10 OF 62 REFERENCES

Characterization of complex networks: A survey of measurements

TLDR
This article presents a survey of measurements capable of expressing the most relevant topological features of complex networks and includes general considerations about complex network characterization, a brief review of the principal models, and the presentation of the main existing measurements.

Statistical Analysis of Network Data

TLDR
This short course will cover a brief overview of the foundations common to the statistical analysis of network data across the disciplines, from a statistical perspective, in the context of topics like network summary and visualization, network sampling, network modeling and inference, and network processes.

Evolution of networks

TLDR
The recent rapid progress in the statistical physics of evolving networks is reviewed, and how growing networks self-organize into scale-free structures is discussed, and the role of the mechanism of preferential linking is investigated.

Latent Space Approaches to Social Network Analysis

TLDR
This work develops a class of models where the probability of a relation between actors depends on the positions of individuals in an unobserved “social space,” and proposes Markov chain Monte Carlo procedures for making inference on latent positions and the effects of observed covariates.

Temporal effects in the growth of networks

TLDR
It is shown that to explain the growth of the citation network by preferential attachment, one has to accept that individual nodes exhibit heterogeneous fitness values that decay with time, which makes the model an apt candidate for modeling a wide range of real systems.

Bootstrap Methods and Their Application

TLDR
Certain selection criteria developed for traditional regression and time series models, when naively applied to certain nonnormal settings, appear to perform at least as well as selection criteria specifically designed for those settings.

Kronecker Graphs: An Approach to Modeling Networks

TLDR
Kronecker graphs naturally obey common network properties and it is rigorously proved that they do so, and KRONFIT, a fast and scalable algorithm for fitting the Kronecker graph generation model to large real networks, is presented.

Statistical mechanics of complex networks

TLDR
A simple model based on these two principles was able to reproduce the power-law degree distribution of real networks, indicating a heterogeneous topology in which the majority of the nodes have a small degree, but there is a significant fraction of highly connected nodes that play an important role in the connectivity of the network.

A coevolving model based on preferential triadic closure for social media networks

TLDR
This paper proposed a coevolving dynamical model for complex networks, in which the evolution is only driven by the local dynamics–the preferential triadic closure, which can reproduce global properties which are qualitatively consistent with the empirical observations.
...