We are developing machine learning (ML) methods to understand how people influence each others’ behavior in social networks. For example, on Twitter, do users influence the content shared or posted by their followers? Methods that can identify such patterns of influence will play a role in studying, e.g., the spread of misinformation on social media sites.

The research intern will be mentored by and work closely with Dhanya Sridhar, a postdoc in the Blei lab. We are extending her initial work on estimating social influence with ML models of social networks. She developed a method that uses observed social network data and applies latent variable models to discover patterns of influence.

There are several important problems that remain as future work, which we will address during this internship. The goals of this internship are two-fold: 1) implement technical extensions that will improve the performance of the method and 2) collect a real-world network dataset to demonstrate applications of our proposed methods.

This is project is NOT accepting applications.

Faculty Advisor

  • Professor: David Blei
  • Department/School: Computer Science
  • Location: Mudd 425
  • The Blei lab focuses on probabilistic machine learning and Bayesian statistics.

Project Timeline

  • Earliest starting date: 10/1/2020
  • End date: 12/15/2020
  • Number of hours per week of research expected during Fall 2020: ~10

Candidate requirements

  • Skill sets: STCS 6701: Foundations of Graphical Models, Proficiency in Python (w/ emphasis on PyTorch), Causal Inference and Approximate Bayesian Inference
  • Student eligibility: freshman, sophomore, junior, senior, master’s
  • International students on F1 or J1 visa: eligible
  • Academic Credit Possible: Yes