## Seminars and Colloquia by Series

Friday, October 19, 2018 - 13:05 , Location: Skiles 005 , , CS, Georgia Tech , , Organizer: He Guo
We investigate whether the standard dimensionality reduction techniques inadvertently produce data representations with different fidelity for two different populations. We show on several real-world datasets, PCA has higher reconstruction error on population A than B (for example, women versus men or lower versus higher-educated individuals). This can happen even when the dataset has similar number of samples from A and B . This motivates our study of dimensionality reduction techniques which maintain similar fidelity for A as B . We give an efficient algorithm for finding a projection which is nearly-optimal with respect to this measure, and evaluate it on several datasets. This is a joint work with&nbsp;Uthaipon Tantipongpipat,&nbsp;Jamie Morgenstern,&nbsp;Mohit Singh, and Santosh Vempala.
Friday, October 12, 2018 - 13:05 , Location: Skiles 005 , , ISyE, Georgia Tech , , Organizer: He Guo
At the heart of most algorithms today there is an optimization engine trying to&nbsp;learn&nbsp;online and provide the best decision, for e.g. rankings of objects, at any time with the partial information observed thus far in time. Often it becomes difficult to find near optimal solutions to many problems due to their inherent&nbsp;combinatorial&nbsp;structure&nbsp;that leads to certain computational bottlenecks. Submodularity is a discrete analogue of convexity and is a key property often exploited in tackling&nbsp;combinatorial&nbsp;optimization problems. In the first part of the talk, we will focus on computational bottlenecks that involve submodular functions: (a) convex function minimization over submodular base polytopes (for e.g. permutahedron) and (b) movement along a line inside submodular base polytopes. We give a conceptually simple and strongly polynomial algorithm Inc-Fix for the former, which is useful in computing Bregman projections in first-order projection-based methods like online mirror descent. For the latter, we will bound the iterations of the discrete Newton method which gives a running time improvement of at least n^6 over the state of the art. This is joint work with Michel Goemans and Patrick Jaillet. In the second part of the talk, we will consider the dual problem of (a), i.e. minimization of composite convex and submodular objectives. We will resolve Bach's conjecture from 2015 about the running time of a popular Kelley's cutting plane variant to minimize these composite objectives. This is joint work with Madeleine Udell and Song Zhou.&nbsp;
Friday, October 5, 2018 - 13:05 , Location: Skiles 005 , , ISyE, Georgia Tech , , Organizer: He Guo
Abstract: Queueing systems are studied in various asymptotic regimes because they are hard to study in general. One popular regime of study is the heavy-traffic regime, when the system is loaded very close to its capacity. Heavy-traffic behavior of queueing systems is traditionally studied using fluid and diffusion limits. In this talk, I will present a recently developed method called the 'Drift Method', which is much simpler, and is based on studying the drift of certain test functions. In addition to exactly characterizing the heavy-traffic behavior, the drift method can be used to obtain lower and upper bounds for all loads. In this talk, I will present the drift method, and its successful application in the context of data center networks to resolve a decade-old conjecture. I will also talk about ongoing work and some open problems. &nbsp; Bio: Siva Theja Maguluri is an Assistant Professor in the School of Industrial and Systems Engineering at Georgia Tech. Before that, he was a Research Staff Member in the Mathematical Sciences Department at IBM T. J. Watson Research Center. He obtained his Ph.D. from the University of Illinois at Urbana-Champaign in Electrical and Computer Engineering where he worked on resource allocation algorithms for cloud computing and wireless networks. Earlier, he received an MS in ECE and an MS in Applied Math from UIUC and a B.Tech in Electrical Engineering from IIT Madras. His research interests include Stochastic Processes, Optimization, Cloud Computing, Data Centers, Resource Allocation and Scheduling Algorithms, Networks, and Game Theory. The current talk is based on a paper that received the best publication in applied probability award, presented by INFORMS Applied probability society.
Friday, September 28, 2018 - 13:15 , Location: Skiles 005 , , Math, Georgia Tech , , Organizer: He Guo
Given a finite partially ordered set (poset) of width $w$, Dilworth's theorem gives an existence and minimality of a chain partition of size $w$. First-Fit is an online algorithm for creating a chain partition of a poset. Given a linear ordering of the points of the poset, $v_1, \cdots, v_n$, First-Fit assigns the point $v_i$ to the first available chain in the chain partition of the points $v_1, \cdots, v_{i-1}$. It is a known fact that First-Fit has unbounded performance over the family of finite posets of width 2. We give a complete characterization of the family of finite posets in which First-Fit performs with subexponential efficiency in terms of width. We will also review what is currently known on the family of posets in which First-Fit performs with polynomial efficiency in terms of width. Joint work with Kevin Milans.
Friday, September 21, 2018 - 13:05 , Location: Skiles 005 , , CS, Georgia Tech , , Organizer: He Guo
As a generalization of many classic problems in combinatorial optimization, submodular optimization has found a wide range of applications in machine learning (e.g., in feature engineering and active learning).&nbsp; For many large-scale optimization problems, we are often concerned with the adaptivity complexity of an algorithm, which quantifies the number of sequential rounds where polynomially-many independent function evaluations can be executed in parallel.&nbsp; While low adaptivity is ideal, it is not sufficient for a (distributed) algorithm to be efficient, since in many practical applications of submodular optimization the number of function evaluations becomes prohibitively expensive.&nbsp; Motivated by such applications, we study the adaptivity and query complexity of adaptive submodular optimization. Our main result is a distributed algorithm for maximizing a monotone submodular function with cardinality constraint $k$ that achieves a $(1-1/e-\varepsilon)$-approximation in expectation.&nbsp; Furthermore, this algorithm runs in $O(\log(n))$ adaptive rounds and makes $O(n)$ calls to the function evaluation oracle in expectation.&nbsp; All three of these guarantees are optimal, and the query complexity is substantially less than in previous works.&nbsp; Finally, to show the generality of our simple algorithm and techniques, we extend our results to the submodular cover problem. Joint work with Vahab Mirrokni and Morteza Zadimoghaddam (<a href="https://arxiv.org/abs/1807.07889">arXiv:1807.07889</a>).
Friday, September 14, 2018 - 13:05 , Location: Skiles 005 , , CS, Georgia Tech , , Organizer: He Guo
&nbsp;We study the dynamic graph connectivity problem in the massively parallel computation model. We give a data structure for maintaining a dynamic undirected graph that handles batches of updates and connectivity queries in constant rounds, as long as the queries fit on a single machine. This assumption corresponds to the gradual buildup of databases over time from sources such as log files and user interactions. Our techniques combine a distributed data structure for Euler Tour (ET) trees, a structural theorem about rapidly contracting graphs via sampling n^{\epsilon} random neighbors, as well as modifications to sketching based dynamic connectivity data structures. Joint work with David Durfee, Janardhan Kulkarni, Richard Peng and Xiaorui Sun.
Friday, April 20, 2018 - 13:05 , Location: Skiles 005 , , CSE, University of Washington , , Organizer: He Guo
We study the $A$-optimal design problem where we are given vectors $v_1,\ldots, v_n\in \R^d$, an integer $k\geq d$, and the goal is to select a set $S$ of $k$ vectors that minimizes the trace of $\left(\sum_{i\in&nbsp; S} v_i v_i^{\top}\right)^{-1}$. Traditionally, the problem is an instance of optimal design of experiments in statistics (\cite{pukelsheim2006optimal}) where each vector corresponds to a linear measurement of an unknown vector and the goal is to pick $k$ of them that minimize the average variance of the error in the maximum likelihood estimate of the vector being measured. The problem also finds applications in sensor placement in wireless networks~(\cite{joshi2009sensor}), sparse least squares regression~(\cite{BoutsidisDM11}), feature selection for $k$-means clustering~(\cite{boutsidis2013deterministic}), and matrix approximation~(\cite{de2007subset,de2011note,avron2013faster}). In this paper, we introduce \emph{proportional volume sampling} to obtain improved approximation algorithms for $A$-optimal design.Given a matrix, proportional volume sampling involves picking a set of columns $S$ of size $k$ with probability proportional to $\mu(S)$ times $\det(\sum_{i \in S}v_i v_i^\top)$ for some measure $\mu$. Our main result is to show the approximability of the $A$-optimal design problem can be reduced to \emph{approximate} independence properties of the measure $\mu$. We appeal to hard-core distributions as candidate distributions $\mu$ that allow us to obtain improved approximation algorithms for the $A$-optimal design. Our results include a $d$-approximation when $k=d$, an $(1+\epsilon)$-approximation when $k=\Omega\left(\frac{d}{\epsilon}+\frac{1}{\epsilon^2}\log\frac{1}{\epsilon}\right)$ and $\frac{k}{k-d+1}$-approximation when repetitions of vectors are allowed in the solution. We also consider generalization of the problem for $k\leq d$ and obtain a $k$-approximation. The last result also implies a restricted invertibility principle for the harmonic mean of singular values.We also show that the $A$-optimal design problem is$\NP$-hard to approximate within a fixed constant when $k=d$.