Mining and Modeling Network Data

Mining and Modeling Network Data is a mini-symposium at the 2018 SIAM Conference on Discrete Mathematics in Denver, Colorado.
Discrete mathematics is at the heart of challenging data-driven problems. Many discrete datasets look like networks, a mathematical formalism for modeling complex systems by interactions between entities. This minisymposium highlights several recent advances in modeling and mining network data with an underlying theme of connecting theory and applications with real data. The theoretical tools are quite diverse and include extremal graph theory, numerical linear algebra, and algorithm design. The applications are equally diverse, drawing from marketing, machine learning, and social network analysis, using data from protein interactions, neural systems, email communications, transportation systems, and more. This minisymposium will showcase the exciting opportunities at the intersection of discrete mathematics and data science.

Session I

  • Austin Benson (Cornell University)
    New perspectives on measuring network clustering
  • David Gleich (Purdue University)
    Hypergraph Kronecker models for networks
  • Rediet Abebe (Cornell University)
    Mitigating Overexposure in Viral Marketing
  • Anthony Bonato (Ryerson University)
    Modeling and mining dynamic competition networks
  • Kathleen Finlinson (University of Colorado Boulder)
    Tuning the activity of neural networks at criticality

Session II

  • Huda Nassar (Purdue University)
    Graph matching via low rank factors
  • Charalampos Tsourakakis (Boston University)
    Risk-Averse Matchings over Uncertain Graph Databases
  • Dane Taylor (SUNY Buffalo)
    Detectability of hierarchical community structure in preprocessed multilayer networks
  • Amir Ghasemian (University of Colorado Boulder)
    Evaluating Overfit and Underfit in Models of Network Community Structure