Johannes Rauh


I am currently interested in the following topics:

  • Complex networks and random graphs

    As part of the VW project on Evoluation of Networks I investigate models for random graphs and random networks. In particular, I am interested in graphons, as limits of finite graphs, and exponential random graphs.

  • Markov bases

    Together with Thomas Kahle I maintain the Markov Bases Database.

  • Conditional Independence models and Robustness

    I'm interested in CI models, their algebraic descriptions and primary decompositions. Some of these models are motivated by studies of robustness, initiated by the VW project on Evoluation of Networks. See this preprint for examples.

  • Algebraic Statistical Models and Exponential Families

    Many statistical models have an algebraic structure that can be exploited in statistics and information theory. For some examples see this article in IJAR and my thesis.

  • Information Geometry: Maximizing the Information Divergence

    Probability distributions that maximize the information divergence from an exponential family have peculiar properties. Studying this maximization problem was the main subject of my thesis. See the group homepage and my thesis for more information.

Workshops and Schools

Publications

Preprints
  • J. Rauh, Optimally approximating exponential families. Preprint: MIS Preprint 73/2011
  • J. Rauh, N. Ay, Robustness and Conditional Independence Ideals. Preprint: arXiv:1110.1338
Journal articles
Conference Contributions
  • N. Ay, G. Montúfar, J. Rauh, Selection Criteria for Neuromanifolds of Stochastic Dynamics. ICCN 2011. Preprint: MIS Preprint 15/2011
  • G. Montúfar, J. Rauh, N. Ay, Expressive Power and Approximation Errors of Restricted Boltzmann Machines. NIPS 2011. Preprint: MIS Preprint 27/2011
Other



Impressum
J.Rauh,  Version vom: 23.2.2010