All Seminars

Show:
Title: Bounded colorings of graphs and hypergraphs
Seminar: Combinatorics
Speaker: Jan Volec of McGill University
Contact: Dwight Duffus, dwight@mathcs.emory.edu
Date: 2017-02-27 at 4:00PM
Venue: W303
Download Flyer
Abstract:
A conjecture of Bollobas and Erdos from 1976 states that any coloring of edges of an n-vertex complete graph such that at each vertex no color appears more than (n/2)-times contains a properly-colored Hamilton cycle. This problem was motivation for the following more general question: Let c be a coloring of E(K_n) where at each vertex, no color appear more than k-times. What properly colored subgraphs does c necessarily contain? In this talk, we will be interested in spanning subgraphs of K_n that have bounded maximum degree or the total number of cherries, i.e., the paths on three vertices. We will also mention similar questions for hypergraphs, as well as analogous problems concerned with rainbow subgraphs in edge colorings of K_n, where the total number of appearances for each color is bounded. One of our main results confirms the following conjecture of Shearer from 1979: If G is an n-vertex graph with O(n) cherries and c is a coloring of E(K_n) such that at each vertex every color appears only constantly many times, then c contains a properly colored copy of G. The talk is based on a joint work with Nina Kamcev and Benny Sudakov.
Title: Uncertainty Quantification and Numerical Analysis: Interactions and Synergies
Seminar: Numerical Analysis and Scientific Computing
Speaker: Daniela Calvetti of Case Western Reserve University
Contact: James Nagy, nagy@mathcs.emory.edu
Date: 2017-02-24 at 1:00PM
Venue: W301
Download Flyer
Abstract:
The computational costs of uncertainty quantification can be challenging, in particular when the problems are large or real time solutions are needed. Numerical methods appropriately modified can turn into powerful and efficient tools for uncertainty quantification. Conversely, state-of-the-art numerical algorithms reinterpreted from the perspective of uncertainty quantification can becomes much more powerful. This presentation will highlight the natural connections between numerical analysis and uncertainty quantification and illustrate the advantages of re-framing classical numerical analysis in a probabilistic setting.
Title: Optical Design from Art to Car Mirrors
Seminar: N/A
Speaker: Sarah Rody of Drexel University
Contact: Bree Ettinger, bree.d.ettinger@emory.edu
Date: 2017-02-24 at 3:00PM
Venue: W201
Download Flyer
Abstract:
In order to design a mirror, we must first decide how to write the problem mathematically. I will start by looking at the historical use of perspective and mirrors in art. Then I will discuss how we can trace individual rays of light to describe how a mirror should work. I will show previous examples of optical design such as a non-reversing mirror and a panoramic mirror. Finally, I will turn to the specific example of a car mirror and show one optical design technique that I use. The standard passenger side mirror on a car has a limited field of view which results in a blind spot. Other mirrors, such as spherical mirrors, reduce the blind spot but distort the image. My goal is to find a construction for a passenger side mirror that reduces the blind spot and but creates less distortion than a spherical mirror. The idea central to our construction is the concept of an eigensurface. In general, if a surface is viewed in a curved mirror, it appears distorted. However, there could exist a surface that appears invariant in a particular curved mirror. I will show how I use this idea of eigensurfaces to find a mirror that could work as a passenger side car mirror.
Title: Good and Bad Reduction of Dynatomic Modular Curves
Seminar: Algebra
Speaker: Andrew Obus of University of Virginia
Contact: David Zureick-Brown, dzb@mathcs.emory.edu
Date: 2017-02-21 at 4:00PM
Venue: W306
Download Flyer
Abstract:
The dynatomic modular curves parameterize one-parameter families of dynamical systems on $P^1$ along with periodic points (or orbits). These are analogous to the standard modular curves parameterizing elliptic curves with torsion points (or subgroups). For the family $x^2 + c$ of quadratic dynamical systems, the corresponding modular curves are smooth in characteristic zero. We give several results about when these curves have good/bad reduction to characteristic $p$, as well as when the reduction is irreducible. These results are motivated by uniform boundedness conjectures in arithmetic dynamics, which will be explained.\\ (This is joint work with John Doyle, Holly Krieger, Rachel Pries, Simon Rubinstein-Salzedo, and Lloyd West.)
Title: Quantum Kostka and the rank on problem for $sl_{2m}$
Seminar: Algebra
Speaker: Natalie Hobson of University of Georgia
Contact: David Zureick-Brown, dzb@mathcs.emory.edu
Date: 2017-02-14 at 4:00PM
Venue: W306
Download Flyer
Abstract:
In this talk we will define and explore an infinite family of vector bundles, known as vector bundles of conformal blocks, on the moduli space $M_{0,n}$ of marked curves. These bundles arise from data associated to a simple Lie algebra. We will show a correspondence (in certain cases) of the rank of these bundles with coefficients in the cohomology of the Grassmannian. This correspondence allows us to use a formula for computing "quantum Kostka" numbers and explicitly characterize families of bundles of rank one by enumerating Young tableaux. We will show these results and illuminate the methods involved.
Title: Torsion subgroups of elliptic curves over quintic and sextic number fields
Seminar: Algebra
Speaker: Maarten Derickx of University of Bayreuth
Contact: David Zureick-Brown, dzb@mathcs.emory.edu
Date: 2017-02-09 at 4:00PM
Venue: W306
Download Flyer
Abstract:
The determination of which finite abelian groups can occur as the torsion subgroup of an elliptic curve over a number field has a long history starting with Barry Mazur who proved that there are exactly 15 groups that can occur as the torsion subgroup of an elliptic curve over the rational numbers. It is a theorem due to Loïc Merel that for every integer d the set of isomorphism classes of groups occurring as the torsion subgroup of a number field of degree d is finite. If a torsion subgroup occurs for a certain degree, then one can also ask for how many distinct pairwise non-isomorphic elliptic curves this happens. The question which torsion groups can occur for infinitely many non-isomorphic elliptic curves of a fixed degree is studied during this talk. The main result is a complete classification of the torsion subgroups that occur infinitely often for degree 5 and 6. This is joint work with Andrew Sutherland and heavily builds on previous joint work with Mark van Hoeij.
Title: The SMURFS Project: Simulation and Modeling for Understanding Resilience and Faults at Scale
Seminar: Computer Science
Speaker: Dorian Arnold of University of New Mexico
Contact: James Lu,
Date: 2017-02-09 at 4:00PM
Venue: W201
Download Flyer
Abstract:
Current HPC research explorations target computer systems with exaflop (10^18 or a quintillion floating point operations per second) capabilities. Such computational power will enable new, important discoveries across all basic science domains. Application resilience is a major challenge to the realization of extreme scale computing systems. The SMURFS Project addresses this challenge by developing methods to improve our predictive understanding of the complex interactions amongst a given application, a given real or hypothetical hardware and software system environment and a given fault-tolerance strategy at extreme scale. Specifically, SMURFS explores: (1) Advanced simulation and modeling capabilities for studying application resilience at scale; (2) Comprehensive, comparative studies of existing and new fault-tolerance strategies; (3) Detailed understandings of how application features interplay with different fault-tolerance strategies and hardware technologies; and (4) Effective prescriptions to guide application developers, hardware architects and system designers to realize efficient, resilient extreme scale capabilities. (This project is a collaboration amongst the University of New Mexico, the University of Tennessee and the Sandia National Labs. It is funded in part by the National Science Foundation.)
Title: Grothendieck Groups and Algebraic K-Theory
Seminar: N/A
Speaker: Juan Villeta-Garcia of University of Illinois Urbana-Champaign
Contact: Bree Ettinger, bree.d.ettinger@emory.edu
Date: 2017-02-06 at 4:00PM
Venue: W303
Download Flyer
Abstract:
Abstract: Mathematics has seen a considerable rise in the interplay of algebra and topology in the last three decades. Algebraic K-Theory, which assigns topological invariants to rings, has profited significantly, with new and profound insights coming out each year. We will give a gentle introduction to Grothendieck Groups (a classical form of Algebraic K-Theory), discuss how they lead to topology, and why trace methods might be useful in computing them.
Title: Privacy-aware task management for mobile crowd sensing
Defense: Dissertation
Speaker: Layla Pournajaf of Emory University
Contact: Layla Pournajaf, layla.pournajaf@emory.edu
Date: 2017-02-03 at 11:00AM
Venue: E408
Download Flyer
Abstract:
Location-aware Mobile crowd sensing (MCS) has numerous applications in a wide range of domains including syndromic surveillance, crime mapping, traffic monitoring, and emergency response. Preserving the privacy of participants in such applications is one of the main challenges in developing effective task management solutions. Moreover, the inherent dynamic environment of MCS characterized by continuous change and uncertain participant movement information pose further challenges for coordination of tasks and participants. Therefore, we propose novel methods to build robust task management frameworks to handle uncertainty and ensure privacy in MCS applications. Our solutions not only increase the disposition of the participants to engage in data collection and sharing activity, but also ultimately lead to more effective MCS applications.
Title: Privacy-Preserving Inference of Social Relationships from Location Data
Seminar: Computer Science
Speaker: Cyrus Shahabi of University of Southern California
Contact: Li Xiong, lxiong@emory.edu
Date: 2017-02-03 at 3:00PM
Venue: W301
Download Flyer
Abstract:
For decades, social scientists have been studying people’s social behaviors by utilizing sparse datasets obtained by observations and surveys. These studies received a major boost in the past decade due to the availability of web data (e.g., social networks, blogs and review web sites). However, due to the nature of the utilized dataset, these studies were confined to behaviors that were observed mostly in the virtual world. Differing from all the earlier work, here, we aim to study social behaviors by observing people’s behaviors in the real world. This is now possible due to the availability of large high-resolution spatiotemporal location data collected by GPS-enabled mobile devices through mobile apps (Google’s Map/Navigation/Search/Chrome, Facebook, Foursquare, WhatsApp, Twitter) or through online services, such as geo-tagged contents (tweets from Twitter, pictures from Instagram, Flickr or Google+ Photo), etc. In particular, we focus on inferring two specific social measures: 1) pair-wise strength -- the strength of social connections between a pair of users, and 2) pair-wise influence - the amount of influence that an individual exerts on another, by utilizing the available high-fidelity location data representing people’s movements. Finally, we argue that due to the sensitivity of location data and user privacy concerns, these inferences cannot be largely carried out on individually contributed data without privacy guarantees. Hence, we discuss open problems in protecting individuals’ location information while enabling these inference analyses.