1703.07317.pdf
6.2 MB
Must read paper on the predictability of #epidemics:
🔹 On the predictability of infectious disease outbreaks
Samuel V. Scarpino and Giovanni Petri
🔹 On the predictability of infectious disease outbreaks
Samuel V. Scarpino and Giovanni Petri
🔸 Do we need to age? How does altruism arise? How do new species form? Explore the basics of evolutionary dynamics:
http://www.necsi.edu/research/evoeco/?platform=hootsuite
http://www.necsi.edu/research/evoeco/?platform=hootsuite
🗞Measurement errors in network data
M. E. J. Newman
🔗 https://arxiv.org/pdf/1703.07376
📌 ABSTRACT
The recent growth in interest in the physics and mathematics of networks has been driven in large part by the increasing availability of data describing the structure of networks ranging from the internet and the web to social and biological networks. It is a surprising feature of many empirical studies, however, that the data are reported without any estimate of their expected accuracy, even though it is clear that most do suffer from measurement error of various kinds. In this paper we develop the theory of measurement error for network data and give an expectation-maximization algorithm for estimating both false positive and false negative rates for edges in observed networks. We give an example application of our methods to social networks determined from proximity data. The methods we describe are general, and could be extended straightforwardly to cover different types of networks or errors.
M. E. J. Newman
🔗 https://arxiv.org/pdf/1703.07376
📌 ABSTRACT
The recent growth in interest in the physics and mathematics of networks has been driven in large part by the increasing availability of data describing the structure of networks ranging from the internet and the web to social and biological networks. It is a surprising feature of many empirical studies, however, that the data are reported without any estimate of their expected accuracy, even though it is clear that most do suffer from measurement error of various kinds. In this paper we develop the theory of measurement error for network data and give an expectation-maximization algorithm for estimating both false positive and false negative rates for edges in observed networks. We give an example application of our methods to social networks determined from proximity data. The methods we describe are general, and could be extended straightforwardly to cover different types of networks or errors.