How reliably can we capture the original network’s properties using a sample? The answer depends on the sampling method. Depending on the problem setting, one may be able to randomly sample nodes or edges; or that may be impossible. One may be able to do Snowball sampling or Respondent driven sampling.
- Statistical properties of sampled networks
- Maximum entropy sampling in complex networks
- The stability of centrality measures when networks are sampled
- On the robustness of centrality measures under conditions of imperfect data
- Online Sampling of High Centrality Individuals in Social Networks
- OMalley2020sampling: The impact of sampling patients on measuring physician patient-sharing networks using Medicare data