Stochastic Step-wise Feature Selection for Exponential Random Graph Models (ERGMs)
2024

New Method for Analyzing Social Networks

Sample size: 9 publication 10 minutes Evidence: high

Author Information

Author(s): Helal El-Zaatari, Fei Yu, Michael R. Kosorok

Primary Institution: University of North Carolina, Chapel Hill, NC, United States of America

Hypothesis

Can a novel methodology for endogenous variable selection in Exponential Random Graph Models (ERGMs) improve the analysis of social networks?

Conclusion

The proposed methodology effectively enhances the analysis of social networks by providing accurate and non-degenerate models.

Supporting Evidence

  • The methodology was tested on nine real-life binary networks.
  • The approach integrates a systematic step-wise feature selection process.
  • It effectively manages the intractable normalizing constants characteristic of ERGMs.
  • The findings contribute to network analysis by offering a robust framework for modeling social networks.

Takeaway

This study created a new way to choose important factors in social networks, helping researchers understand how people and organizations connect better.

Methodology

The study introduces a step-wise feature selection process for ERGMs, addressing challenges like degeneracy and computational complexity.

Limitations

The model space can become inflated when including directed endogenous variables, complicating the selection process.

Digital Object Identifier (DOI)

10.1371/journal.pone.0314557

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication