Using bootstrap for statistical inference on random graphs
arXiv:1402.3647 · doi:10.1002/cjs.11271
Abstract
In this paper, we propose new nonparametric approach to network inference that may be viewed as a fusion of block sampling procedures for temporally and spatially dependent processes with the classical network methodology. We develop estimation and uncertainty quantification procedures for network mean degree using a "patchwork" sample and nonparametric bootstrap, under the assumption of unknown degree distribution. We investigate asymptotic properties of the proposed patchwork bootstrap procedure and present cross-validation methodology for selecting an optimal patch size. We validate the new patchwork bootstrap on simulated networks with short and long tailed mean degree distributions, and revisit the Erdos collaboration data to illustrate the proposed methodology.
arXiv commentThe paper has been withdrawn by the authors: a general revision of methodology is needed
References in corpus (11)
- Finding and evaluating community structure in networks
- Power-law distributions in empirical data
- Random graphs with arbitrary degree distributions and their applications
- The spread of epidemic disease on networks
- Evolution of the social network of scientific collaborations
- Modeling social networks from sampled data
- Traceroute sampling makes random graphs appear to have power law degree distributions
- Accuracy and Scaling Phenomena in Internet Mapping
- The method of moments and degree distributions for network models
- Subsampling bootstrap of count features of networks
- Non-parametric resampling of random walks for spectral network clustering