# MSCS Seminar Calendar

Monday May 2, 2016
Computer Science Seminar
Learning Communities in the Presence of Errors
Aravindan Vijayaraghavan (Northwestern)
3:00 PM in SEO 612
The Stochastic Block Model or the Planted Partition Model is the most widely used probabilistic model for community detection and clustering graphs in various fields, including machine learning, statistics, and social sciences. Many existing algorithms successfully learn the communities or clusters when the data is drawn exactly according to the model, but they do not work well in the presence of errors. In this talk, I will address the following question: Can we design robust polynomial time algorithms for learning probabilistic models for community detection that work in the presence of adversarial modeling errors? I will present robust algorithms for (partially) recovering communities or clusters in graphs drawn from the Stochastic Block Model, in the presence of modeling errors or noise. These algorithms allow two types of adversarial errors: edge outlier errors (where an adversary can corrupt an arbitrary \epsilon fraction of the edges), and any amount of Feige-Kilian or monotone errors. Mossel, Neeman and Sly (STOC 2015) posed an open question about whether an almost exact recovery is possible when the adversary is allowed to add o(n) edges. Our work answers this question affirmatively even in the case of k>2 communities.
Finally, I will describe how our algorithms recover the clusters, even when the modeling error is captured using Kullback-Leibler (KL) divergence: these algorithms work when the instances come from any distribution of graphs that is \epsilon.m close to the Stochastic Block Model in the KL divergence (this result also handles adversarial errors).
This is based on joint work with Konstantin Makarchev and Yury Makarychev.
Wednesday May 4, 2016
Number Theory Seminar
A heuristic for boundedness of elliptic curves
Jennifer Park (University of Michigan)
1:00 PM in SEO 1227
I will discuss a heuristic that predicts that the ranks of all but finitely many elliptic curves defined over Q are bounded above by 21. This is joint work with Bjorn Poonen, John Voight, and Melanie Matchett Wood.
Friday May 6, 2016
Number Theory Seminar
A p-adic strengthening of the Manin-Mumford conjecture
Vlad Serban (Northwestern)
11:00 AM in SEO 427
Let $G$ be an abelian variety or a product of multiplicative groups $\mathbb{G}_m^n$ and let $C$ be an embedded curve. The Manin-Mumford conjecture (a theorem by work of Lang, Raynaud et al.) states that only finitely many torsion points of $G$ can lie on $C$ unless $C$ is in fact a subgroup of $G$. I will show how these purely algebraic statements extend to suitable analytic functions on open $p$-adic unit poly-disks. These disks occur naturally as weight spaces parametrizing families of $p$-adic automorphic forms for $GL(2)$ over a number field $F$. When $F=\mathbb{Q}$, the "Hida families" in question play a crucial role in the study of modular forms. When $F$ is imaginary quadratic, I will explain how our results imply that Bianchi modular forms are sparse in these $p$-adic families.
Monday May 9, 2016
Thesis Defense
Moduli Spaces of Sheaves on $\mathbb{P}^1 \times \mathbb{P}^1$
Tim Ryan (UIC)
11:00 AM in SEO 636
In this talk, after reviewing the basic properties of moduli spaces of sheaves on $\mathbb{P}^1 \times \mathbb{P}^1$, I will show that they are $\mathbb{Q}$-factorial Mori Dream Spaces and explain a method for computing their effective cones. My method is based on the generalized Beilinson spectral sequence, Bridgeland stability, and moduli spaces of Kronecker modules.
Monday August 29, 2016
Geometry, Topology and Dynamics Seminar
Random Grids in G-spaces
Jayadev S. Athreya (University of Washington)
3:00 PM in SEO 636
We show how to define a general notion of a random grid, how it generalizes the notion of random Euclidean lattices and random affine lattices, and describe how it gives a natural notion of a random hyperbolic lattice. This is joint work with Gregory Margulis and Yair Minsky.
We will be going for a seminar dinner. Email schapos@uic.edu if you'd like to join.

Analysis and Applied Mathematics Seminar
TBA
Jiahong Wu (Oklahoma State)
4:00 PM in SEO 636
Wednesday September 7, 2016
Statistics Seminar
TBA
Andrey Sarantsev (University of California, Santa Barbara)
4:00 PM in SEO 636
TBA
Wednesday September 21, 2016
Statistics Seminar
TBA
Yan Chang (Discover)
4:00 PM in SEO 636
Monday September 26, 2016
Geometry, Topology and Dynamics Seminar
TBA
Jordan Watts (University of Colorado Boulder)
3:00 PM in SEO 636
TBA
Monday October 3, 2016
Geometry, Topology and Dynamics Seminar
TBA
Lucas Branco (University of Oxford)
3:00 PM in SEO 636
TBA
We will be going for a seminar dinner. Email schapos@uic.edu if you'd like to join.
UIC LAS MSCS > seminars > seminar calendar