Friday, March 17, 2017 - 15:00 , Location: Skiles 006 , Rick Durett , Duke University , firstname.lastname@example.org , Organizer: Megan Bernstein
In the latent voter model, which models the spread of a technology through a social network, individuals who have just changed their choice have a latent period, which is exponential with rate λ during which they will not buy a new device. We study site and edge versions of this model on random graphs generated by a configuration model in which the degrees d(x) have 3 ≤ d(x) ≤ M. We show that if the number of vertices n → ∞ and log n << λn << n then the latent voter model has a quasi-stationary state in which each opinion has probability ≈ 1/2 and persists in this state for a time that is ≥ nm for any m <∞. Thus, even a very small latent period drastically changes the behavior of the voter model.
Friday, March 17, 2017 - 14:00 , Location: Skiles 006 , John Etnyre , Georgia Tech , Organizer: John Etnyre
This will be a 1.5 hour (maybe slightly longer) seminar.
Following up on the previous series of talks we will show how to construct Lagrangian Floer homology and discuss it properties.
Series: ACO Student Seminar
Optimization problems arising in decentralized multi-agent systems have gained significant attention in the context of cyber-physical, communication, power, and robotic networks combined with privacy preservation, distributed data mining and processing issues. The distributed nature of the problems is inherent due to partial knowledge of the problem data (i.e., a portion of the cost function or a subset of the constraints is known to different entities in the system), which necessitates costly communications among neighboring agents. In this talk, we present a new class of decentralized first-order methods for nonsmooth and stochastic optimization problems which can significantly reduce the number of inter-node communications. Our major contribution is the development of decentralized communication sliding methods, which can skip inter-node communications while agents solve the primal subproblems iteratively through linearizations of their local objective functions.This is a joint work with Guanghui (George) Lan and Yi Zhou.
Series: Algebra Seminar
Error-correcting decoding is generalized to multivariate sparse polynomial and rational function interpolation from evaluations that can be numerically inaccurate and where several evaluations can have severe errors (``outliers''). Our multivariate polynomial and rational function interpolation algorithm combines Zippel's symbolic sparse polynomial interpolation technique [Ph.D. Thesis MIT 1979] with the numeric algorithm by Kaltofen, Yang, and Zhi [Proc. SNC 2007], and removes outliers (``cleans up data'') by techniques from the Welch/Berlekamp decoder for Reed-Solomon codes. Our algorithms can build a sparse function model from a number of evaluations that is linear in the sparsity of the model, assuming that there are a constant number of ouliers and that the function probes can be randomly chosen.
Series: School of Mathematics Colloquium
We present algorithms for performing sparse univariate polynomial interpolation with errors in the evaluations of the polynomial. Our interpolation algorithms use as a substep an algorithm that originally is by R. Prony from the French Revolution (Year III, 1795) for interpolating exponential sums and which is rediscovered to decode digital error correcting BCH codes over finite fields (1960). Since Prony's algorithm is quite simple, we will give a complete description, as an alternative for Lagrange/Newton interpolation for sparse polynomials. When very few errors in the evaluations are permitted, multiple sparse interpolants are possible over finite fields or the complex numbers, but not over the real numbers. The problem is then a simple example of list-decoding in the sense of Guruswami-Sudan. Finally, we present a connection to the Erdoes-Turan Conjecture (Szemeredi's Theorem). This is joint work with Clement Pernet, Univ. Grenoble.
Series: Graph Theory Seminar
For a graph G, the Colin de Verdière graph parameter mu(G) is the maximum corank of any matrix in a certain family of generalized adjacency matrices of G. Given a non-negative integer t, the family of graphs with mu(G) <= t is minor-closed and therefore has some nice properties. For example, a graph G is planar if and only if mu(G) <= 3. Colin de Verdière conjectured that the chromatic number chi(G) of a graph satisfies chi(G) <= mu(G)+1. For graphs with mu(G) <= 3 this is the Four Color Theorem. We conjecture that if G has at least t vertices and mu(G) <= t, then |E(G)| <= t|V(G)| - (t+1 choose 2). For planar graphs this says |E(G)| <= 3|V(G)|-6. If this conjecture is true, then chi(G) <= 2mu(G). We prove the conjectured edge upper bound for certain classes of graphs: graphs with mu(G) small, graphs with mu(G) close to |V(G)|, chordal graphs, and the complements of chordal graphs.
Series: Stochastics Seminar
Estimation of the covariance matrix has attracted significant attention of the statistical research community over the years, partially due to important applications such as Principal Component Analysis. However, frequently used empirical covariance estimator (and its modifications) is very sensitive to outliers, or ``atypical’’ points in the sample. As P. Huber wrote in 1964, “...This raises a question which could have been asked already by Gauss, but which was, as far as I know, only raised a few years ago (notably by Tukey): what happens if the true distribution deviates slightly from the assumed normal one? As is now well known, the sample mean then may have a catastrophically bad performance…” Motivated by Tukey's question, we develop a new estimator of the (element-wise) mean of a random matrix, which includes covariance estimation problem as a special case. Assuming that the entries of a matrix possess only finite second moment, this new estimator admits sub-Gaussian or sub-exponential concentration around the unknown mean in the operator norm. We will present extensions of our approach to matrix-valued U-statistics, as well as applications such as the matrix completion problem. Part of the talk will be based on a joint work with Xiaohan Wei.
Thursday, March 16, 2017 - 11:00 , Location: Skiles 005 , Rick Durrett , Duke University , email@example.com , Organizer: Megan Bernstein
The use of evolutionary game theory biology dates to work of Maynard-Smith who used it to explain why most fights between animals were of the limited war type. Nowak and collaborators have shown that a spatial distribution of players can explain the existence of altruism, which would die out in a homogeneously mixing population. For the last twenty years, evolutionary games have been used to model cancer. In applications to ecology and cancer, the system is not homogeneously mixing so it is important to understand how space changes the outcome of these games. Over the last several years we have developed a theory for understanding the behavior of evolutionary games in the weak selection limit. We will illustrate this theory by discussing a number of examples. The most recent work was done in collaboration with a high school student so the talk should be accessible to a broad audience.