Seminars and Colloquia Schedule

Discrete Mathematical Biology Working Seminar

Series
Other Talks
Time
Monday, March 26, 2012 - 11:00 for 1 hour (actually 50 minutes)
Location
Skiles 114
Speaker
Will Perkins – Georgia Tech
A discussion of the paper "Complete suboptimal folding of RNA and the stability of secondary structures" by Wuchty et al (1999).

Got symmetry? Here is how you slice it

Series
CDSNS Colloquium
Time
Monday, March 26, 2012 - 11:00 for 1 hour (actually 50 minutes)
Location
Skiles 006
Speaker
Predrag Cvitanovic – Georgia Tech, Physics
With recent advances in experimental imaging, computational methods, and dynamics insights it is now possible to start charting out the terra incognita explored by turbulence in strongly nonlinear classical field theories, such as fluid flows. In presence of continuous symmetries these solutions sweep out 2- and higher-dimensional manifolds (group orbits) of physically equivalent states, interconnected by a web of still higher-dimensional stable/unstable manifolds, all embedded in the PDE infinite-dimensional state spaces. In order to chart such invariant manifolds, one must first quotient the symmetries, i.e. replace the dynamics on M by an equivalent, symmetry reduced flow on M/G, in which each group orbit of symmetry-related states is replaced by a single representative.Happy news: The problem has been solved often, first by Jacobi (1846), then by Hilbert and Weyl (1921), then by Cartan (1924), then by [...], then in this week's arXiv [...]. Turns out, it's not as easy as it looks.Still, every unhappy family is unhappy in its own way: The Hilbert's solution (invariant polynomial bases) is useless. The way we do this in quantum field theory (gauge fixing) is not right either. Currently only the "method of slices" does the job: it slices the state space by a set of hyperplanes in such a way that each group orbit manifold of symmetry-equivalent points is represented by a single point, but as slices are only local, tangent charts, an atlas comprised from a set of charts is needed to capture the flow globally. Lots of work and not a pretty sight (Nature does not like symmetries), but one is rewarded by much deeper insights into turbulent dynamics; without this atlas you will not get anywhere.This is not a fluid dynamics talk. If you care about atomic, nuclear or celestial physics, general relativity or quantum field theory you might be interested and perhaps help us do this better.You can take part in this seminar from wherever you are by clicking onevo.caltech.edu/evoNext/koala.jnlp?meeting=M2MvMB2M2IDsDs9I9lDM92

Flame-pressure interactions and stretched laminar flame velocities: implicit simulation methods with realistic chemistry

Series
Math Physics Seminar
Time
Monday, March 26, 2012 - 12:00 for 1 hour (actually 50 minutes)
Location
Skiles 006
Speaker
Nadeem Malik – King Fahd University of Petroleum and Minerals
An implicit method [1, 2], TARDIS (Transient Advection Reaction Diffusion Implicit Simulations), has been developed that successfully couples the compressible flow to the comprehensive chemistry and multi-component transport properties. TARDIS has been demonstrated in application to two fundamental combustion problems of great interest. First, TARDIS was used to investigate stretched laminar flame velocities in eight flame configurations: outwardly and inwardly propagating H2/air and CH4/air in cylindrical and spherical geometries. Fractional power laws are observed between the velocity deficit and the flame curvature Second, the response of transient outwardly propagating premixed H2/air and CH4/air flames subjected to joint pressure and equivalence ratio oscillations were investigated. A fuller version of the abstract can be obtained from http://www.math.gatech.edu/~rll6/malik_abstract-Apr-2012.docx [1] Malik, N.A. and Lindstedt, R.P. The response of transient inhomogeneous flames to pressure fluctuations and stretch: planar and outwardly propagating hydrogen/air flames. Combust. Sci. Tech. 82(9), 2010. [2] Malik, N. A. “Fractional powers laws in stretched flame velocities in finite thickness flames: a numerical study using realistic chemistry”. Under review, (2012). [3] Markstein, G.H. Non-steady Flame Propagation. Pergamon Press, 1964. [4] Weis,M., Zarzalis, N., and Suntz, R. Experimental study of markstein number effects on laminar flamelet velocity in turbulent premixed flames. Combust. Flame, 154:671--691, 2008.

New Numerical Linear Algebra Techniques for Brownian Simulation of Macromolecules

Series
Applied and Computational Mathematics Seminar
Time
Monday, March 26, 2012 - 14:00 for 1 hour (actually 50 minutes)
Location
Skiles 006
Speaker
Edmond Chow – School of Computational Science and Engineering, Georgia Institute of Technology
Brownian dynamics (BD) is a computational technique for simulating the motions of molecules interacting through hydrodynamic and non-hydrodynamic forces. BD simulations are the main tool used in computational biology for studying diffusion-controlled cellular processes. This talk presents several new numerical linear algebra techniques to accelerate large BD simulations, and related Stokesian dynamics (SD) simulations. These techniques include: 1) a preconditioned Lanczos process for computing Brownian vectors from a distribution with given covariance, 2) low-rank approximations to the hydrodynamic tensor suitable for large-scale problems, and 3) a reformulation of the computations to approximate solutions to multiple time steps simultaneously, allowing the efficient use of data parallel hardware on modern computer architectures.

Truncated Toeplitz operators

Series
Analysis Seminar
Time
Monday, March 26, 2012 - 14:00 for 1 hour (actually 50 minutes)
Location
Skiles 114
Speaker
Dan Timotin – Indiana University and Mathematical Institute of Romania
Truncated Toeplitz operators, introduced in full generality by Sarason a few years ago, are compressions of multiplication operators on H^2 to subspaces invariant to the adjoint of the shift. The talk will survey this newly developing area, presenting several of the basic results and highlighting some intriguing open questions.

Curve complex translation lengths

Series
Geometry Topology Seminar
Time
Monday, March 26, 2012 - 14:05 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Vaibhav Gadre – Harvard University
The curve complex C(S) of a closed orientable surface S of genusg is an infinite graph with vertices isotopy classes of essential simpleclosed curves on S with two vertices adjacent by an edge if the curves canbe isotoped to be disjoint. By a celebrated theorem of Masur-Minsky, thecurve complex is Gromov hyperbolic. Moreover, a pseudo-Anosov map f of Sacts on C(S) as a hyperbolic isometry with "north-south" dynamics and aninvariant quasi-axis. One can define an asymptotic translation length for fon C(S). In joint work with Chia-yen Tsai, we prove bounds on the minimalpseudo-Anosov asymptotic translation lengths on C(S) . We shall alsooutline related interesting results and questions.

Hamilton-Jacobi equations on metric spaces and transport entropy inequalities

Series
PDE Seminar
Time
Tuesday, March 27, 2012 - 15:05 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Cyril Roberto – University of Paris, Nanterre
We will prove an Hopf-Lax-Oleinik formula for the solutions of some Hamilton-Jacobi equations on a general metric space. Then, we will present some consequences: in particular the equivalence of the log-Sobolev inequality and the hypercontractivity property of theHamilton-Jacobi "semi-group", (and if time allows) that Talagrand’s transport-entropy inequalities in metric space are characterizedin terms of log-Sobolev inequalities restricted to the class of c-convex functions (based on a paper in collaboration with N. Gozlan and P.M. Samson).

From birational invariants to elections

Series
Research Horizons Seminar
Time
Wednesday, March 28, 2012 - 12:05 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Anton Leykin – Georgia Tech
This talk will traverse several topics in singularity theory, algebraic analysis, complex analysis, algebraic geometry, and statistics. I will outline effective methods to compute the log canonical threshold, a birational invariant of an algebraic variety, as well as its potential statistical applications.

Augmenting undirected node-connectivity by one - Part II

Series
Graph Theory Seminar
Time
Thursday, March 29, 2012 - 12:05 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Laszlo Vegh – College of Computing, Georgia Tech
In the node-connectivity augmentation problem, we want to add a minimum number of new edges to an undirected graph to make it k-node-connected. The complexity of this question is still open, although the analogous questions of both directed and undirected edge-connectivity and directed node-connectivity augmentation are known to be polynomially solvable. I present a min-max formula and a polynomial time algorithm for the special case when the input graph is already (k-1)-connected. The formula has been conjectured by Frank and Jordan in 1994. In the first lecture, I presented previous results on the other connectivity augmentation variants. In the second part, I shall present my min-max formula and the main ideas of the proof.

From Statistical to Game-Theoretic Learning

Series
Stochastics Seminar
Time
Thursday, March 29, 2012 - 15:05 for 1 hour (actually 50 minutes)
Location
skyles 006
Speaker
Alexander Rakhlin – University of Pennsylvania, The Wharton School
The study of prediction within the realm of Statistical Learning Theory is intertwined with the study of the supremum of an empirical process. The supremum can be analyzed with classical tools:Vapnik-Chervonenkis and scale-sensitive combinatorial dimensions, covering and packing numbers, and Rademacher averages. Consistency of empirical risk minimization is known to be closely related to theuniform Law of Large Numbers for function classes.In contrast to the i.i.d. scenario, in the sequential prediction framework we are faced with an individual sequence of data on which weplace no probabilistic assumptions. The problem of universal prediction of such deterministic sequences has been studied withinStatistics, Information Theory, Game Theory, and Computer Science. However, general tools for analysis have been lacking, and mostresults have been obtained on a case-by-case basis.In this talk, we show that the study of sequential prediction is closely related to the study of the supremum of a certain dyadic martingale process on trees. We develop analogues of the Rademacher complexity, covering numbers and scale-sensitive dimensions, which canbe seen as temporal generalizations of the classical results. The complexities we define also ensure uniform convergence for non-i.i.d. data, extending the Glivenko-Cantelli type results. Analogues of local Rademacher complexities can be employed for obtaining fast rates anddeveloping adaptive procedures. Our understanding of the inherent complexity of sequential prediction is complemented by a recipe that can be used for developing new algorithms.

Approximating CSPs with Global Cardinality Constraints Using SDP Hierarchies

Series
ACO Student Seminar
Time
Friday, March 30, 2012 - 13:00 for 1 hour (actually 50 minutes)
Location
Skiles 005
Speaker
Ning Tan – School of Math, Georgia Tech –
In a celebrated result, Raghavendra [Rag08] showed that, assuming Unique Game Conjecture, every Max-CSP problem has a sharp approximation threshold that matches the integrality gap of a natural SDP relaxation. Raghavendra and Steurer [RS09] also gave a simple and unified framework for optimally approximating all the Max-CSPs.In this work, we consider the problem of approximating CSPs with global cardinality constraints. For example, Max-Cut is a boolean CSP where the input is a graph $G = (V,E)$ and the goal is to find a cut $S \cup \bar S = V$ that maximizes the number of crossing edges, $|E(S,\bar S)|$. The Max-Bisection problem is a variant of Max-Cut with an additional global constraint that each side of the cut has exactly half the vertices, i.e., $|S| = |V|/2$. Several other natural optimization problems like Small Set Expansion, Min Bisection and approximating Graph Expansion can be formulated as CSPs with global constraints.In this talk, I will introduce a general approach towards approximating CSPs with global constraints using SDP hierarchies. To demonstrate the approach, I will present an improved algorithm for Max-Bisection problem that achieves the following:- Given an instance of Max-Bisection with value $1-\epsilon$, the algorithm finds a bisection with value at least $1-O(\sqrt{\epsilon})$ with running time $O(n^{poly(1/\eps)})$. This approximation is near-optimal (up to constant factors in $O()$) under the Unique Games Conjecture.- Using computer-assisted proof, we show that the same algorithm also achieves a 0.85-approximation for Max-Bisection, improving on the previous bound of 0.70 (note that it is UGC-hard to approximate better than 0.878 factor). As an attempt to prove matching hardness result, we show a generic conversion from SDP integrality gap to dictatorship test for any CSP with global cardinality constraints. The talk is based on joint work with Prasad Raghavendra.

Plane fields on 3-manifolds I

Series
Geometry Topology Working Seminar
Time
Friday, March 30, 2012 - 14:00 for 2 hours
Location
Skiles 006
Speaker
John Etnyre – Ga Tech

Note this is a 2 hour talk

In this series of talks I will discuss various special plane fields on 3-manifold. Specifically we will consider folaitions and contact structures and the relationship between them. We will begin by sketching a proof of Eliashberg and Thurston's famous theorem from the 1990's that says any sufficiently smooth foliation can be approximated by a contact structure. In the remaining talks I will discuss ongoing research that sharpens our understanding of the relation between foliations and contact structures.