When: 1 April, 1-2 pm

Where: B705

Abstract

Connections among people in the form of (binary) social networks have proven important in the study of social support, disease spread, and a host of other domains. In many of these settings it is infeasible to collect exhaustive community-level networks, even when communities are relatively small. Traditional sampling techniques are not capable of capturing any of the network structure as real-world networks are sparse – in an n-size SRS of a large N population only a fraction n/N of the sampled nodes’ ties will be observed. A number of network-based sampling approaches have been proposed that aim to retain the structure of the network. Here we consider statistical inference for network processed from partially observed networks in the context of data collected in the field. The network models concern either tie-formation or processes on networks.

Find out more about Johan Koskinen's Research: https://findanexpert.unimelb.edu.au/profile/136290-johan-koskinen