Where: Kirwan Hall 1311

Speaker: Maria Cameron and Kasso Okoudjou (UMCP) - http://www.math.umd.edu/~mariakc/rit-analysis-of-complex.html

Where: Kirwan Hall 1311

Speaker: Maria Cameron (UMCP, Math) - http://www.math.umd.edu/~mariakc/

Abstract: I will discuss the construction and analysis of continuous-time discrete-space Markov chains (stochastic networks) modeling the aggregation of interacting particles into clusters.

I will consider two types of particle interaction: via the Lennard-Jones potential which is an adequate model for cooled rare gas atoms, and via a short-range or sticky potential which is an adequate model for micron size styrofoam balls immersed in the water. I will provide physical motivations for considering these problems, explain difficulties arising in the building of the networks, and outline challenges in their analysis and perspectives.

Where: Kirwan Hall 1311

Speaker: Maria Cameron (UMCP Math) - http://www.math.umd.edu/~mariakc/rit-analysis-of-complex.html

Abstract: It will be an introductory talk explaining what are regulatory networks and why they are appealing modeling tools. I will discuss the modeling of a budding yeast cell cycle by means of a regulatory network. The dynamics of the resulting network is consistent with the cell cycle. Then I will establish which edges in the regulatory network are redundant and which are essential by removing one edge in a time and checking whether the resulting network still follows the cell cycle.

Where: Kirwan Hall 1311

Speaker: Kasso Okoudjou (UMCP, Mathematics) - https://www.math.umd.edu/~okoudjou/

Where: Kirwan Hall 1311

Speaker: Kasso Okoudjou (UMCP, Mathematics) - https://www.math.umd.edu/~okoudjou/

Where: Kirwan Hall 1311

Speaker: Matt Yancey () -

Abstract: Based on recent results characterizing the curvature of several real-world networks, there have been recent attempts at determining what, if anything, the curvature implies. In this talk, the version of curvature that we will investigate is Gromov's relaxation of Buneman's 4-point condition to define when a group is hyperbolic. The proposed consequence of hyperbolicity we investigate is congestion: that there exists a node $u$ and fixed $\epsilon > 0$ such that for randomly uniformly selected endpoints $v,w$, the geodesic from $v$ to $w$ contains $u$ has probability at least $\epsilon$.

There exist trivial/degenerate examples of non-congested hyperbolic networks. We will present a series of results demonstrating that hyperbolicity combined with other natural conditions imply congestion.

Where: Kirwan Hall 1311

Speaker: Kasso Okoudjou (UMCP Math) - https://www.math.umd.edu/~kasso