Event Detail

Event Type: 
Monday, January 22, 2018 - 16:00 to 17:00
GLSN 200

Speaker Info

Cornell University

Preferential attachment is a mechanism for modeling power-law behavior of the degree distributions in directed social networks. We consider methods for fitting a 5-parameter linear preferential model to network data under two data scenarios. In the case where full history of the network formation is given, we derive the maximum likelihood estimator of the parameters and show that they are strongly consistent and asymptotically normal. In the case where only a {single-time} snapshot of the network is available, we propose an estimation method which combines method of moments with an approximation to the likelihood. The resulting estimator is also strongly consistent and performs well compared to the MLE estimator. We illustrate both estimation procedures using simulated data, and explore the usage of this model in a real data example. We also present a semi-parametric method to model heavy-tailed features of the degree distributions of the network using ideas from extreme value theory and compare advantages (robustness against model error) and disadvantages of asymptotic methods (reduced efficiency if the model is correct).

This is a joint work with Tiandong Wang, Cornell; Richard Davis & Phyllis Wan, Columbia