BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CMSA - ECPv6.15.18//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:CMSA
X-ORIGINAL-URL:https://cmsa.fas.harvard.edu
X-WR-CALDESC:Events for CMSA
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20140309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20141102T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20150308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20151101T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20160313T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20161106T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20170312T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20171105T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20180311T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20181104T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20190310T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20191103T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20180124T090000
DTEND;TZID=America/New_York:20180125T170000
DTSTAMP:20260427T225110
CREATED:20230717T173945Z
LAST-MODIFIED:20250305T214037Z
UID:10000042-1516784400-1516899600@cmsa.fas.harvard.edu
SUMMARY:Blockchain Conference
DESCRIPTION:On January 24-25\, 2019 the Center of Mathematical Sciences will be hosting a conference on distributed-ledger (blockchain) technology. The conference is intended to cover a broad range of topics\, from abstract mathematical aspects (cryptography\, game theory\, graph theory\, theoretical computer science) to concrete applications (in accounting\, government\, economics\, finance\, management\, medicine). The talks will take place in Science Center\, Hall D. \nhttps://youtu.be/FyKCCutxMYo \nPhotos\n \nSpeakers: \n\nJoseph Abadi\, Princeton University\nBenedikt Bunz\, Stanford University\nJake Cacciapaglia\, Nebula Genomics/Harvard Medical School\nEduardo Castello\, Massachusetts Institute of Technology\nAlisa DiCaprio\, R3\nZhiguo He\, University of Chicago\nSteven Kou\, Boston University\nAnne Lafarre\, Tilburg University\nJacob Leshno\, University of Chicago\nBruce Schneier\, Harvard Kennedy School\nDavid Schwartz\, Ripple\nElaine Shi\, Cornell University/Thunder Research\nHong Wan\, NCSU
URL:https://cmsa.fas.harvard.edu/event/blockchain-conference/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Conference,Event
ATTACH;FMTTYPE=image/jpeg:https://cmsa.fas.harvard.edu/media/Blockchain-Final-scaled.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20171010T170000
DTEND;TZID=America/New_York:20171010T180000
DTSTAMP:20260427T225110
CREATED:20230717T173349Z
LAST-MODIFIED:20250328T150724Z
UID:10000038-1507654800-1507658400@cmsa.fas.harvard.edu
SUMMARY:2017 Ding Shum Lecture
DESCRIPTION:Leslie Valiant will be giving the inaugural talk of the Ding Shum Lectures on Tuesday\, October 10 at 5:00 pm in Science Center Hall D\, Cambridge\, MA. \nLearning as a Theory of Everything \nAbstract: We start from the hypothesis that all the information that resides in living organisms was initially acquired either through learning by an individual or through evolution. Then any unified theory of evolution and learning should be able to characterize the capabilities that humans and other living organisms can possess or acquire. Characterizing these capabilities would tell us about the nature of humans\, and would also inform us about feasible targets for automation. With this purpose we review some background in the mathematical theory of learning. We go on to explain how Darwinian evolution can be formulated as a form of learning. We observe that our current mathematical understanding of learning is incomplete in certain important directions\, and conclude by indicating one direction in which further progress would likely enable broader phenomena of intelligence and cognition to be realized than is possible at present. \n 
URL:https://cmsa.fas.harvard.edu/event/2017-ding-shum-lecture/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Ding Shum Lecture,Event,Public Lecture,Special Lectures
ATTACH;FMTTYPE=image/png:https://cmsa.fas.harvard.edu/media/Ding-Shum-lecture-3.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20171002T091500
DTEND;TZID=America/New_York:20171002T173000
DTSTAMP:20260427T225110
CREATED:20230717T172938Z
LAST-MODIFIED:20250328T150846Z
UID:10000036-1506935700-1506965400@cmsa.fas.harvard.edu
SUMMARY:The 2017 Charles River Lectures
DESCRIPTION:Charles River with Bench at Sunset\nJointly organized by Harvard University\, Massachusetts Institute of Technology\, and Microsoft Research New England\, the Charles River Lectures on Probability and Related Topics is a one-day event for the benefit of the greater Boston area mathematics community. \nThe 2017 lectures will take place 9:15am – 5:30pm on Monday\, October 2 at Harvard University  in the Harvard Science Center. \n\n\n\n*************************************************** \nUPDATED LOCATION\nHarvard University\nHarvard Science Center (Halls C & E)\n1 Oxford Street\, Cambridge\, MA 02138 (Map)\nMonday\, October 2\, 2017\n9:15 AM – 5:30 PM\n************************************************** \nPlease note that registration has closed. \nSpeakers:\n\nPaul Bourgade (Courant Institute\, NYU)\nMassimiliano Gubinelli (University of Bonn)\nAndrea Montanari (Stanford University)\nRoman Vershynin (University of California\, Irvine)\nOfer Zeitouni (Weizmann Institute)\n\nAgenda:\nIn Harvard Science Center Hall C: \n8:45 am – 9:15 am: Coffee/light breakfast \n9:15 am – 10:15 am: Ofer Zeitouni \nTitle: Noise stability of the spectrum of large matrices \nAbstract: The spectrum of large non-normal matrices is notoriously sensitive to perturbations\, as the example of nilpotent matrices shows. Remarkably\, the spectrum of these matrices perturbed by polynomially (in the dimension) vanishing additive noise is remarkably stable. I will describe some results and the beginning of a theory. \nThe talk is based on joint work with Anirban Basak and Elliot Paquette\, and earlier works with Feldheim\, Guionnet\, Paquette and Wood.\n\n10:20 am – 11:20 am: Andrea Montanari \nTitle: Algorithms for estimating low-rank matrices  \nAbstract: Many interesting problems in statistics can be formulated as follows. The signal of interest is a large low-rank matrix with additional structure\, and we are given a single noisy view of this matrix. We would like to estimate the low rank signal by taking into account optimally the signal structure. I will discuss two types of efficient estimation procedures based on message-passing algorithms and semidefinite programming relaxations\, with an emphasis on asymptotically exact results. \n11:20 am – 11:45 am: Break \n11:45 am – 12:45 pm: Paul Bourgade \nTitle: Random matrices\, the Riemann zeta function and trees \nAbstract: Fyodorov\, Hiary & Keating have conjectured that the maximum of the characteristic polynomial of random unitary matrices behaves like extremes of log-correlated Gaussian fields. This allowed them to predict the typical size of local maxima of the Riemann zeta function along the critical axis. I will first explain the origins of this conjecture\, and then outline the proof for the leading order of the maximum\, for unitary matrices and the zeta function. This talk is based on joint works with Arguin\, Belius\, Radziwill and Soundararajan. \n1:00 pm – 2:30 pm: Lunch \nIn Harvard Science Center Hall E: \n2:45 pm – 3:45 pm: Roman Vershynin \nTitle: Deviations of random matrices and applications \nAbstract: Uniform laws of large numbers provide theoretical foundations for statistical learning theory. This lecture will focus on quantitative uniform laws of large numbers for random matrices. A range of illustrations will be given in high dimensional geometry and data science. \n3:45 pm – 4:15 pm: Break \n4:15 pm – 5:15 pm: Massimiliano Gubinelli \nTitle: Weak universality and Singular SPDEs \nAbstract: Mesoscopic fluctuations of microscopic (discrete or continuous) dynamics can be described in terms of nonlinear stochastic partial differential equations which are universal: they depend on very few details of the microscopic model. This universality comes at a price: due to the extreme irregular nature of the random field sample paths\, these equations turn out to not be well-posed in any classical analytic sense. I will review recent progress in the mathematical understanding of such singular equations and of their (weak) universality and their relation with the Wilsonian renormalisation group framework of theoretical physics. \nOrganizers:\n Alexei Borodin\, Henry Cohn\, Vadim Gorin\, Elchanan Mossel\, Philippe Rigollet\, Scott Sheffield\, and H.T. Yau
URL:https://cmsa.fas.harvard.edu/event/the-2017-charles-river-lectures/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Event,Public Lecture,Special Lectures
ATTACH;FMTTYPE=image/jpeg:https://cmsa.fas.harvard.edu/media/Charles-River-Lectures-2017-pdf.jpeg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20170818T154700
DTEND;TZID=America/New_York:20170819T154700
DTSTAMP:20260427T225110
CREATED:20230717T172600Z
LAST-MODIFIED:20250328T144515Z
UID:10000034-1503071220-1503157620@cmsa.fas.harvard.edu
SUMMARY:2017 Big Data Conference
DESCRIPTION:The Center of Mathematical Sciences and Applications will be hosting a conference on Big Data from August 18 – 19\, 2017\, in Hall D of the Science Center at Harvard University.\nThe Big Data Conference features many speakers from the Harvard community as well as scholars from across the globe\, with talks focusing on computer science\, statistics\, math and physics\, and economics. This is the third conference on Big Data the Center will host as part of our annual events\, and is co-organized by Richard Freeman\, Scott Kominers\, Jun Liu\, Horng-Tzer Yau and Shing-Tung Yau. \nConfirmed Speakers: \n\nMohammad Akbarpour\, Stanford University\nAlbert-László Barabási\, Northeastern University\nNoureddine El Karoui\, University of California\, Berkeley\nRavi Jagadeesan\, Harvard University\nLucas Janson\, Harvard University\nTracy Ke\, University of Chicago\nTze Leung Lai\, Stanford University\nAnnie Liang\, University of Pennsylvania\nMarena Lin\, Harvard University\nNikhil Naik\, Harvard University\nAlex Peysakhovich\, Facebook\nNatesh Pillai\, Harvard University\nJann Spiess\, Harvard University\nBradly Stadie\, Open AI\, University of California\, Berkeley\nZak Stone\, Google\nHau-Tieng Wu\, University of Toronto\nSifan Zhou\, Xiamen University\n\n  \nFollowing the conference\, there will be a two-day workshop from August 20-21. The workshop is organized by Scott Kominers\, and will feature: \n\nJörn Boehnke\, Harvard University\nNikhil Naik\, Harvard University\nBradly Stadie\, Open AI\, University of California\, Berkeley\n\n  \nConference Schedule \nA PDF version of the schedule below can also be downloaded here. \nAugust 18\, Friday (Full day)\n\n\n\nTime\nSpeaker\nTopic\n\n\n8:30 am – 9:00 am\n\nBreakfast\n\n\n9:00 am – 9:40 am\nMohammad Akbarpour \nVideo\nTitle: Information aggregation in overlapping generations and the emergence of experts \nAbstract: We study a model of social learning with “overlapping generations”\, where agents meet others and share data about an underlying state over time. We examine under what conditions the society will produce individuals with precise knowledge about the state of the world. There are two information sharing regimes in our model: Under the full information sharing technology\, individuals exchange the information about their point estimates of an underlying state\, as well as their sources (or the precision of their signals) and update their beliefs by taking a weighted average. Under the limited information sharing technology\, agents only observe the information about the point estimates of those they meet\, and update their beliefs by taking a weighted average\, where weights can depend on the sequence of meetings\, as well as the labels. Our main result shows that\, unlike most social learning settings\, using such linear learning rules do not guide the society (or even a fraction of its members) to learn the truth\, and having access to\, and exploiting knowledge of the precision of a source signal are essential for efficient social learning (joint with Amin Saberi & Ali Shameli).\n\n\n9:40 am – 10:20 am\nLucas Janson \nVideo\nTitle: Model-Free Knockoffs For High-Dimensional Controlled Variable Selection \nAbstract: Many contemporary large-scale applications involve building interpretable models linking a large set of potential covariates to a response in a nonlinear fashion\, such as when the response is binary. Although this modeling problem has been extensively studied\, it remains unclear how to effectively control the fraction of false discoveries even in high-dimensional logistic regression\, not to mention general high-dimensional nonlinear models. To address such a practical problem\, we propose a new framework of model-free knockoffs\, which reads from a different perspective the knockoff procedure (Barber and Candès\, 2015) originally designed for controlling the false discovery rate in linear models. The key innovation of our method is to construct knockoff variables probabilistically instead of geometrically. This enables model-free knockoffs to deal with arbitrary (and unknown) conditional models and any dimensions\, including when the dimensionality p exceeds the sample size n\, while the original knockoffs procedure is constrained to homoscedastic linear models with n greater than or equal to p. Our approach requires the design matrix be random (independent and identically distributed rows) with a covariate distribution that is known\, although we show our procedure to be robust to unknown/estimated distributions. As we require no knowledge/assumptions about the conditional distribution of the response\, we effectively shift the burden of knowledge from the response to the covariates\, in contrast to the canonical model-based approach which assumes a parametric model for the response but very little about the covariates. To our knowledge\, no other procedure solves the controlled variable selection problem in such generality\, but in the restricted settings where competitors exist\, we demonstrate the superior power of knockoffs through simulations. Finally\, we apply our procedure to data from a case-control study of Crohn’s disease in the United Kingdom\, making twice as many discoveries as the original analysis of the same data. \nSlides\n\n\n10:20 am – 10:50 am\n\nBreak\n\n\n10:50 pm – 11:30 pm\nNoureddine El Karoui \nVideo\nTitle: Random matrices and high-dimensional statistics: beyond covariance matrices \nAbstract: Random matrices have played a central role in understanding very important statistical methods linked to covariance matrices (such as Principal Components Analysis\, Canonical Correlation Analysis etc…) for several decades. In this talk\, I’ll show that one can adopt a random-matrix-inspired point of view to understand the performance of other widely used tools in statistics\, such as M-estimators\, and very common methods such as the bootstrap. I will focus on the high-dimensional case\, which captures well the situation of “moderately” difficult statistical problems\, arguably one of the most relevant in practice. In this setting\, I will show that random matrix ideas help upend conventional theoretical thinking (for instance about maximum likelihood methods) and highlight very serious practical problems with resampling methods.\n\n\n11:30 am – 12:10 pm\nNikhil Naik \nVideo\nTitle: Understanding Urban Change with Computer Vision and Street-level Imagery \nAbstract: Which neighborhoods experience physical improvements? In this work\, we introduce a computer vision method to measure changes in the physical appearances of neighborhoods from time-series street-level imagery. We connect changes in the physical appearance of five US cities with economic and demographic data and find three factors that predict neighborhood improvement. First\, neighborhoods that are densely populated by college-educated adults are more likely to experience physical improvements. Second\, neighborhoods with better initial appearances experience\, on average\, larger positive improvements. Third\, neighborhood improvement correlates positively with physical proximity to the central business district and to other physically attractive neighborhoods. Together\, our results illustrate the value of using computer vision methods and street-level imagery to understand the physical dynamics of cities. \n(Joint work with Edward L. Glaeser\, Cesar A. Hidalgo\, Scott Duke Kominers\, and Ramesh Raskar.)\n\n\n12:10 pm – 12:25 pm\nVideo #1 \nVideo #2\nData Science Lightning Talks\n\n\n12:25 pm – 1:30 pm\n\nLunch\n\n\n1:30 pm – 2:10 pm\nTracy Ke \nVideo\nTitle: A new SVD approach to optimal topic estimation \nAbstract: In the probabilistic topic models\, the quantity of interest—a low-rank matrix consisting of topic vectors—is hidden in the text corpus matrix\, masked by noise\, and Singular Value Decomposition (SVD) is a potentially useful tool for learning such a low-rank matrix. However\, the connection between this low-rank matrix and the singular vectors of the text corpus matrix are usually complicated and hard to spell out\, so how to use SVD for learning topic models faces challenges. \nWe overcome the challenge by revealing a surprising insight: there is a low-dimensional simplex structure which can be viewed as a bridge between the low-rank matrix of interest and the SVD of the text corpus matrix\, and which allows us to conveniently reconstruct the former using the latter. Such an insight motivates a new SVD-based approach to learning topic models. \nFor asymptotic analysis\, we show that under a popular topic model (Hofmann\, 1999)\, the convergence rate of the l1-error of our method matches that of the minimax lower bound\, up to a multi-logarithmic term. In showing these results\, we have derived new element-wise bounds on the singular vectors and several large deviation bounds for weakly dependent multinomial data. Our results on the convergence rate and asymptotical minimaxity are new. We have applied our method to two data sets\, Associated Process (AP) and Statistics Literature Abstract (SLA)\, with encouraging results. In particular\, there is a clear simplex structure associated with the SVD of the data matrices\, which largely validates our discovery.\n\n\n2:10 pm – 2:50 pm\nAlbert-László Barabási \nVideo\nTitle: Taming Complexity: From Network Science to Controlling Networks \nAbstract: The ultimate proof of our understanding of biological or technological systems is reflected in our ability to control them. While control theory offers mathematical tools to steer engineered and natural systems towards a desired state\, we lack a framework to control complex self-organized systems. Here we explore the controllability of an arbitrary complex network\, identifying the set of driver nodes whose time-dependent control can guide the system’s entire dynamics. We apply these tools to several real networks\, unveiling how the network topology determines its controllability. Virtually all technological and biological networks must be able to control their internal processes. Given that\, issues related to control deeply shape the topology and the vulnerability of real systems. Consequently unveiling the control principles of real networks\, the goal of our research\, forces us to address series of fundamental questions pertaining to our understanding of complex systems. \n \n\n\n2:50 pm – 3:20 pm\n\nBreak\n\n\n3:20 pm – 4:00 pm\nMarena Lin \nVideo\nTitle: Optimizing climate variables for human impact studies \nAbstract: Estimates of the relationship between climate variability and socio-economic outcomes are often limited by the spatial resolution of the data. As studies aim to generalize the connection between climate and socio-economic outcomes across countries\, the best available socio-economic data is at the national level (e.g. food production quantities\, the incidence of warfare\, averages of crime incidence\, gender birth ratios). While these statistics may be trusted from government censuses\, the appropriate metric for the corresponding climate or weather for a given year in a country is less obvious. For example\, how do we estimate the temperatures in a country relevant to national food production and therefore food security? We demonstrate that high-resolution spatiotemporal satellite data for vegetation can be used to estimate the weather variables that may be most relevant to food security and related socio-economic outcomes. In particular\, satellite proxies for vegetation over the African continent reflect the seasonal movement of the Intertropical Convergence Zone\, a band of intense convection and rainfall. We also show that agricultural sensitivity to climate variability differs significantly between countries. This work is an example of the ways in which in-situ and satellite-based observations are invaluable to both estimates of future climate variability and to continued monitoring of the earth-human system. We discuss the current state of these records and potential challenges to their continuity.\n\n\n4:00 pm – 4:40 pm\nAlex Peysakhovich\n Title: Building a cooperator \nAbstract: A major goal of modern AI is to construct agents that can perform complex tasks. Much of this work deals with single agent decision problems. However\, agents are rarely alone in the world. In this talk I will discuss how to combine ideas from deep reinforcement learning and game theory to construct artificial agents that can communicate\, collaborate and cooperate in productive positive sum interactions.\n\n\n4:40 pm – 5:20 pm\nTze Leung Lai \nVideo\nTitle: Gradient boosting: Its role in big data analytics\, underlying mathematical theory\, and recent refinements \nAbstract: We begin with a review of the history of gradient boosting\, dating back to the LMS algorithm of Widrow and Hoff in 1960 and culminating in Freund and Schapire’s AdaBoost and Friedman’s gradient boosting and stochastic gradient boosting algorithms in the period 1999-2002 that heralded the big data era. The role played by gradient boosting in big data analytics\, particularly with respect to deep learning\, is then discussed. We also present some recent work on the mathematical theory of gradient boosting\, which has led to some refinements that greatly improves the convergence properties and prediction performance of the methodology.\n\n\n\nAugust 19\, Saturday (Full day)\n\n\n\nTime\nSpeaker\nTopic\n\n\n8:30 am – 9:00 am\n\nBreakfast\n\n\n9:00 am – 9:40 am\nNatesh Pillai \nVideo\nTitle: Accelerating MCMC algorithms for Computationally Intensive Models via Local Approximations \nAbstract: We construct a new framework for accelerating Markov chain Monte Carlo in posterior sampling problems where standard methods are limited by the computational cost of the likelihood\, or of numerical models embedded therein. Our approach introduces local approximations of these models into the Metropolis–Hastings kernel\, borrowing ideas from deterministic approximation theory\, optimization\, and experimental design. Previous efforts at integrating approximate models into inference typically sacrifice either the sampler’s exactness or efficiency; our work seeks to address these limitations by exploiting useful convergence characteristics of local approximations. We prove the ergodicity of our approximate Markov chain\, showing that it samples asymptotically from the exact posterior distribution of interest. We describe variations of the algorithm that employ either local polynomial approximations or local Gaussian process regressors. Our theoretical results reinforce the key observation underlying this article: when the likelihood has some local regularity\, the number of model evaluations per Markov chain Monte Carlo (MCMC) step can be greatly reduced without biasing the Monte Carlo average. Numerical experiments demonstrate multiple order-of-magnitude reductions in the number of forward model evaluations used in representative ordinary differential equation (ODE) and partial differential equation (PDE) inference problems\, with both synthetic and real data.\n\n\n9:40 am – 10:20 am\nRavi Jagadeesan \nVideo\nTitle: Designs for estimating the treatment effect in networks with interference \nAbstract: In this paper we introduce new\, easily implementable designs for drawing causal inference from randomized experiments on networks with interference. Inspired by the idea of matching in observational studies\, we introduce the notion of considering a treatment assignment as a quasi-coloring” on a graph. Our idea of a perfect quasi-coloring strives to match every treated unit on a given network with a distinct control unit that has identical number of treated and control neighbors. For a wide range of interference functions encountered in applications\, we show both by theory and simulations that the classical Neymanian estimator for the direct effect has desirable properties for our designs. This further extends to settings where homophily is present in addition to interference.\n\n\n10:20 am – 10:50 am\n\nBreak\n\n\n10:50 am – 11:30 am\nAnnie Liang \nVideo\nTitle: The Theory is Predictive\, but is it Complete? An Application to Human Generation of Randomness \nAbstract: When we test a theory using data\, it is common to focus on correctness: do the predictions of the theory match what we see in the data? But we also care about completeness: how much of the predictable variation in the data is captured by the theory? This question is difficult to answer\, because in general we do not know how much “predictable variation” there is in the problem. In this paper\, we consider approaches motivated by machine learning algorithms as a means of constructing a benchmark for the best attainable level of prediction.  We illustrate our methods on the task of predicting human-generated random sequences. Relative to a theoretical machine learning algorithm benchmark\, we find that existing behavioral models explain roughly 15 percent of the predictable variation in this problem. This fraction is robust across several variations on the problem. We also consider a version of this approach for analyzing field data from domains in which human perception and generation of randomness has been used as a conceptual framework; these include sequential decision-making and repeated zero-sum games. In these domains\, our framework for testing the completeness of theories provides a way of assessing their effectiveness over different contexts; we find that despite some differences\, the existing theories are fairly stable across our field domains in their performance relative to the benchmark. Overall\, our results indicate that (i) there is a significant amount of structure in this problem that existing models have yet to capture and (ii) there are rich domains in which machine learning may provide a viable approach to testing completeness (joint with Jon Kleinberg and Sendhil Mullainathan).\n\n\n11:30 am – 12:10 pm\nZak Stone \nVideo\nTitle: TensorFlow: Machine Learning for Everyone \nAbstract: We’ve witnessed extraordinary breakthroughs in machine learning over the past several years. What kinds of things are possible now that weren’t possible before? How are open-source platforms like TensorFlow and hardware platforms like GPUs and Cloud TPUs accelerating machine learning progress? If these tools are new to you\, how should you get started? In this session\, you’ll hear about all of this and more from Zak Stone\, the Product Manager for TensorFlow on the Google Brain team.\n\n\n12:10 pm – 1:30 pm\n\nLunch\n\n\n1:30 pm – 2:10 pm\nJann Spiess \nVideo\nTitle: (Machine) Learning to Control in Experiments \nAbstract: Machine learning focuses on high-quality prediction rather than on (unbiased) parameter estimation\, limiting its direct use in typical program evaluation applications. Still\, many estimation tasks have implicit prediction components. In this talk\, I discuss accounting for controls in treatment effect estimation as a prediction problem. In a canonical linear regression framework with high-dimensional controls\, I argue that OLS is dominated by a natural shrinkage estimator even for unbiased estimation when treatment is random; suggest a generalization that relaxes some parametric assumptions; and contrast my results with that for another implicit prediction problem\, namely the first stage of an instrumental variables regression.\n\n\n2:10 pm – 2:50 pm\nBradly Stadie\nTitle: Learning to Learn Quickly: One-Shot Imitation and Meta Learning \nAbstract: Many reinforcement learning algorithms are bottlenecked by data collection costs and the brittleness of their solutions when faced with novel scenarios.\nWe will discuss two techniques for overcoming these shortcomings. In one-shot imitation\, we train a module that encodes a single demonstration of a desired behavior into a vector containing the essence of the demo. This vector can subsequently be utilized to recover the demonstrated behavior. In meta-learning\, we optimize a policy under the objective of learning to learn new tasks quickly. We show meta-learning methods can be accelerated with the use of auxiliary objectives. Results are presented on grid worlds\, robotics tasks\, and video game playing tasks.\n\n\n2:50 pm – 3:20 pm\n\nBreak\n\n\n3:20 pm – 4:00 pm\nHau-Tieng Wu \nVideo\nTitle: When Medical Challenges Meet Modern Data Science \nAbstract: Adaptive acquisition of correct features from massive datasets is at the core of modern data analysis. One particular interest in medicine is the extraction of hidden dynamics from a single observed time series composed of multiple oscillatory signals\, which could be viewed as a single-channel blind source separation problem. The mathematical and statistical problems are made challenging by the structure of the signal which consists of non-sinusoidal oscillations with time varying amplitude/frequency\, and by the heteroscedastic nature of the noise. In this talk\, I will discuss recent progress in solving this kind of problem by combining the cepstrum-based nonlinear time-frequency analysis and manifold learning technique. A particular solution will be given along with its theoretical properties. I will also discuss the application of this method to two medical problems – (1) the extraction of a fetal ECG signal from a single lead maternal abdominal ECG signal; (2) the simultaneous extraction of the instantaneous heart/respiratory rate from a PPG signal during exercise; (3) (optional depending on time) an application to atrial fibrillation signals. If time permits\, the clinical trial results will be discussed.\n\n\n4:00 pm – 4:40 pm\nSifan Zhou \nVideo\nTitle: Citing People Like Me: Homophily\, Knowledge Spillovers\, and Continuing a Career in Science \nAbstract: Forward citation is widely used to measure the scientific merits of articles. This research studies millions of journal article citation records in life sciences from MEDLINE and finds that authors of the same gender\, the same ethnicity\, sharing common collaborators\, working in the same institution\, or being geographically close are more likely (and quickly) to cite each other than predicted by their proportion among authors working on the same research topics. This phenomenon reveals how social and geographic distances influence the quantity and speed of knowledge spillovers. Given the importance of forward citations in academic evaluation system\, citation homophily potentially put authors from minority group at a disadvantage. I then show how it influences scientists’ chances to survive in the academia and continue publishing. Based on joint work with Richard Freeman.\n\n\n\n  \nTo view photos and video interviews from the conference\, please visit the CMSA blog. \n\n \n\n  \n\n\n\nBig Data\,CMSA\,Harvard\,Math\nEvents\,Past Events
URL:https://cmsa.fas.harvard.edu/event/2017-big-data-conference-aug-18-19/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Big Data Conference,Conference,Event
ATTACH;FMTTYPE=image/png:https://cmsa.fas.harvard.edu/media/Big-Data-2017_2.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20170428T090000
DTEND;TZID=America/New_York:20170502T170000
DTSTAMP:20260427T225110
CREATED:20230717T175015Z
LAST-MODIFIED:20250305T215930Z
UID:10000030-1493370000-1493744400@cmsa.fas.harvard.edu
SUMMARY:JDG 2017 Conference
DESCRIPTION:In celebration of the Journal of Differential Geometry’s 50th anniversary\, the Harvard Math Department will be hosting the Tenth Conference on Geometry and Topology (JDG 2017) from April 28 – May 2\, 2017. \nConfirmed Speakers \n\nMina Aganagic\, UC Berkeley\nDenis Auroux\, UC Berkeley\nCaucher Birkar\, University of Cambridge\nHuai-Dong Cao\, Lehigh University\nTristan Collins\, Harvard University\nCamillo De Lellis\, ETH Zurich\nJean-Pierre Demailly\, Grenoble Alpes University\nSimon Donaldson\, Stony Brook University\nDan Freed\, University of Texas at Austin\nKenji Fukaya\, Stony Brook University\nDavid Gabai\, Princeton University\nLarry Guth\, Massachusetts Institute of Technology\nRichard Hamilton\, Columbia University\nYujiro Kawamata\, University of Tokyo\nFrances Kirwan\, Oxford University\nBlaine Lawson\, Stony Brook University\nJun Li\, Stanford University\nSi Li\, Tsinghua University\nBong Lian\, Brandeis University\nChiu-Chu Melissa Liu\, Columbia University\nCiprian Manolescu\, University of California\, Los Angeles\nFernando Marques\, Princeton University\nWilliam Meeks\, University of Massachusetts Amherst\nWilliam Minicozzi\, Massachusetts Institute of Technology\nJohn Pardon\, Princeton University\nDuong Phong\, Columbia University\nAlena Pirutka\, Courant Institute of New York University\nRichard Schoen\, University of California\, Irvine\nArtan Sheshmani\, QGM Aarhus University/Harvard University\nCliff Taubes\, Harvard University\nCumrun Vafa\, Harvard University\nMu-Tao Wang\, Columbia University\nShing-Tung Yau\, Harvard University\nSteve Zelditch\, Northwestern University\n\n* This event is co-sponsored by Lehigh University and partially supported by the National Science Foundation.
URL:https://cmsa.fas.harvard.edu/event/jdg-2017-conference-april-28-may-2-2017/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Conference,Event
ATTACH;FMTTYPE=image/jpeg:https://cmsa.fas.harvard.edu/media/JDG-2017-scaled.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20160822T090000
DTEND;TZID=America/New_York:20160823T163000
DTSTAMP:20260427T225110
CREATED:20230717T171959Z
LAST-MODIFIED:20250328T144123Z
UID:10000017-1471856400-1471969800@cmsa.fas.harvard.edu
SUMMARY:2016 Big Data Conference & Workshop
DESCRIPTION:! LOCATION CHANGE: The conference will be in Science Center Hall C on Tuesday\, Aug.23\, 2016.\nThe Center of Mathematical Sciences and Applications will be hosting a workshop on Big Data from August 12 – 21\, 2016 followed by a two-day conference on Big Data from August 22 – 23\, 2016. \nBig Data Conference features many speakers from the Harvard Community as well as many scholars from across the globe\, with talks focusing on computer science\, statistics\, math and physics\, and economics. This is the second conference on Big Data the Center will host as part of our annual events. The 2015 conference was a huge success. \nThe conference will be hosted at Harvard Science Center Hall A (Monday\, Aug.22) & Hall C (Tuesday\, Aug.23): 1 Oxford Street\, Cambridge\, MA 02138. \nThe 2016 Big Data conference is sponsored by the Center of Mathematical Sciences and Applications at Harvard University and the Alfred P. Sloan Foundation. \nConference Speakers:\n\nJörn Boehnke\, Harvard CMSA\nJoan Bruna\, UC Berkeley [Video]\nTamara Broderick\, MIT [Video]\nJustin Chen\, MIT [Video]\nYiling Chen\, Harvard University [Video]\nAmir Farbin\, UT Arlington [Video]\nDoug Finkbeiner\, Harvard University [Video]\nAndrew Gelman\, Columbia University [Video]\nNina Holden\, MIT [Video]\nElchanan Mossel\, MIT\nAlex Peysakhovich\, Facebook\nAlexander Rakhlin\, University of Pennsylvania [Video]\nNeal Wadhwa\, MIT [Video]\nJun Yin\, University of Wisconsin\nHarry Zhou\, Yale University [Video]\n\nPlease click Conference Program for a downloadable schedule with talk abstracts.\nConference Schedule:\n\n\n\nAugust 22 – Day 1\n\n\n8:30am\nBreakfast\n\n\n8:55am\nOpening remarks\n\n\n9:00am – 9:50am\nYiling Chen\, “Machine Learning with Strategic Data Sources” [Video]\n\n\n9:50am – 10:40am\nAndrew Gelman\, “Taking Bayesian Inference Seriously” [Video]\n\n\n10:40am – 11:10am\nBreak\n\n\n11:10am – 12:00pm\nHarrison Zhou\, “A General Framework for Bayes Structured Linear Models” [Video]\n\n\n12:00pm – 1:30pm\nLunch\n\n\n1:30pm – 2:20pm\nDouglas Finkbeiner\, “Mapping the Milky Way in 3D with star colors” [Video]\n\n\n2:20pm – 3:10pm\nNina Holden\, “Sparse exchangeable graphs and their limits” [Video]\n\n\n3:10pm – 3:40pm\nBreak\n\n\n3:40pm – 4:30pm\nAlex Peysakhovich\, “How social science methods inform personalization on Facebook News Feed” [Video]\n\n\n4:30pm – 5:20pm\nAmir Farbin\, “Deep Learning in High Energy Physics” [Video]\n\n\n\n\n\nAugust 23 – Day 2\n\n\n8:45am\nBreakfast\n\n\n9:00am – 9:50am\nJoan Bruna Estrach\, “Addressing Computational and Statistical Gaps with Deep Networks” [Video]\n\n\n9:50am – 10:40am\nJustin Chen & Neal Wadhwa\, “Smaller Than the Eye Can See: Big Engineering from Tiny Motions in Video” [Video]\n\n\n10:40am – 11:10am\nBreak\n\n\n11:10am – 12:00pm\nAlexander Rakhlin\, “How to Predict When Estimation is Hard: Algorithms for Learning on Graphs” [Video]\n\n\n12:00pm – 1:30pm\nLunch\n\n\n1:30pm – 2:20pm\nTamara Broderick\, “Fast Quantification of Uncertainty and Robustness with Variational Bayes” [Video]\n\n\n2:20pm – 3:10pm\nElchanan Mossel\, “Phylogenetic Reconstruction – a Rigorous Model of Deep Learning”\n\n\n3:10pm – 3:40pm\nBreak\n\n\n3:40pm – 4:30pm\nJörn Boehnke\, “Amazon’s Price and Sales-rank Data: What can one billion prices on 150 thousand products tell us about the economy?”\n\n\n\nWorkshop Participants:\nRichard Freeman’s Group: \n\nSen Chai\, ESSEC\nBrock Mendel\, Harvard University\nRaviv Muriciano-Goroff\, Stanford University\nSifan Zhou\, CMSA\n\nScott Kominer’s Group: \n\nBradly Stadie\, UC Berkeley\nNeal Wadhwa\, MIT [Video]\nJustin Chen\n\nChristopher Rogan’s Group: \n\nAmir Farbin\, UT Arlington [Video]\nPaul Jackson\, University of Adelaide\n\nFor more information about the workshops\, please reach out directly to the individual group leaders. \n* This event is sponsored by CMSA Harvard University and the Alfred P. Sloan Foundation. \n 
URL:https://cmsa.fas.harvard.edu/event/2016-big-data-conference-workshop/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Big Data Conference,Conference,Event,Workshop
ATTACH;FMTTYPE=image/png:https://cmsa.fas.harvard.edu/media/Big-Data_2016_2-1-2.png
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20150824T084500
DTEND;TZID=America/New_York:20150826T160000
DTSTAMP:20260427T225110
CREATED:20230717T180044Z
LAST-MODIFIED:20250304T180628Z
UID:10000013-1440405900-1440604800@cmsa.fas.harvard.edu
SUMMARY:2015 Conference on Big Data
DESCRIPTION:The Center of Mathematical Sciences and Applications will be having a conference on Big Data August 24-26\, 2015\, in Science Center Hall B at Harvard University.  This conference will feature many speakers from the Harvard Community as well as many scholars from across the globe\, with talks focusing on computer science\, statistics\, math and physics\, and economics.\n\n \nMonday\, August 24 \n\n\n\nTime\nSpeaker\nTitle\n\n\n8:45am\nMeet and Greet\n\n\n\n9:00am\nSendhil Mullainathan\nPrediction Problems in Social Science: Applications of Machine Learning to Policy and Behavioral Economics\n\n\n9:45am\nMike Luca\nDesigning Disclosure for the Digital Age\n\n\n10:30\nBreak\n\n\n\n10:45\nJianqing Fan\nBig Data Big Assumption: Spurious discoveries and endogeneity\n\n\n11:30am\nDaniel Goroff\nPrivacy and Reproducibility in Data Science\n\n\n12:15pm\nBreak for Lunch\n\n\n\n2:00pm\nRyan Adams\nExact Markov Chain Monte Carlo with Large Data\n\n\n2:45pm\nDavid Dunson\nScalable Bayes: Simple algorithms with guarantees\n\n\n3:30pm\nBreak\n\n\n\n3:45pm\nMichael Jordan\nComputational thinking\, inferential thinking and Big Data\n\n\n4:30pm\nJoel Tropp\nApplied Random Matrix Theory\n\n\n5:15pm\nDavid Woodruff\nInput Sparsity and Hardness for Robust Subspace Approximation\n\n\n\nTuesday\, August 25 \n\n\n\nTime\nSpeaker\nTitle\n\n\n8:45am\nMeet and Greet\n\n\n\n9:00am\nGunnar Carlsson\nPersistent homology for qualitative analysis and feature generation\n\n\n9:45am\nAndrea Montanari\nSemidefinite Programming Relaxations for Graph and Matrix Estimation: Algorithms and Phase Transitions\n\n\n10:30am\nBreak\n\n\n\n10:45am\nSusan Athey\nMachine Learning and Causal Inference for Policy Evaluation\n\n\n11:30am\nDenis Nekipelov\nRobust Empirical Evaluation of Large Competitive Markets\n\n\n12:15pm\nBreak for Lunch\n\n\n\n2:00pm\nLucy Colwell\nUsing evolutionary sequence variation to make inferences about protein structure and function: Modeling with Random Matrix Theory\n\n\n2:45pm\nSimona Cocco\nInverse Statistical Physics approaches for the modeling of protein families\n\n\n3:30pm\nBreak\n\n\n\n3:45pm\nRemi Monasson\nInference of top components of correlation matrices with prior informations\n\n\n4:30pm\nSayan Mukherjee\nRandom walks on simplicial complexes and higher order notions of spectral clustering\n\n\n\n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \n  \nA Banquet from 7:00 – 8:30pm will follow Tuesday’s talks. This event is by invitation only. \n Wednesday\, August 26  \n\n\n\nTime\nSpeaker\nTitle\n\n\n8:45am\nMeet and Greet\n\n\n\n9:00am\nAnkur Moitra\nBeyond Matrix Completion\n\n\n9:45am\nFlorent Krzakala\nOptimal compressed sensing with spatial coupling and message passing\n\n\n10:30am\nBreak\n\n\n\n10:45am\nPiotr Indyk\nFast Algorithms for Structured Sparsity\n\n\n11:30am\nGuido Imbens\nExact p-values for network inference\n\n\n12:15pm\nBreak for lunch\n\n\n\n2:00pm\nEdo Airoldi\nSome fundamental ideas for causal inference on large networks\n\n\n2:45pm\nRonitt Rubinfeld\nSomething for almost nothing: sublinear time approximation algorithms\n\n\n3:30pm\nBreak\n\n\n\n3:45pm\nLenka Zdeborova\nClustering of sparse networks:  Phase transitions and optimal algorithms\n\n\n4:30pm\nJelani Nelson\nDimensionality reductions via sparse matrices
URL:https://cmsa.fas.harvard.edu/event/conference-on-big-data-august-24-26-2015/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Big Data Conference,Conference,Event
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20150817T140000
DTEND;TZID=America/New_York:20150817T170000
DTSTAMP:20260427T225110
CREATED:20230904T081317Z
LAST-MODIFIED:20250305T171843Z
UID:10000050-1439820000-1439830800@cmsa.fas.harvard.edu
SUMMARY:GAMES ON HETEROGENEOUS GRAPHS
DESCRIPTION:A major challenge in evolutionary biology is to understand how spatial population structure affects the evolution of social behaviors such as\ncooperation. This question can be investigated mathematically by studying evolutionary processes on graphs. Individuals occupy vertices and interact with neighbors according to a matrix game. Births and deaths occur stochastically according to an update rule. Previously\, full mathematical results have only been obtained for graphs with strong symmetry properties. Our group is working to extend these results to certain classes of asymmetric graphs\, using tools such as random walk theory and harmonic analysis. \n \n\n\n  \nHere is a list of the scholars participating in this program. \n\n\n\n\nName\n\n\n\n\nShing-Tung Yau\n\n\nMartin Nowak\n\n\nBen Adlam\n\n\nBen Allen\n\n\nYu-Ting Chen\n\n\nAn Huang\n\n\nGabor Lippner
URL:https://cmsa.fas.harvard.edu/event/games-on-heterogeneous-graphs/
LOCATION:Harvard Science Center\, 1 Oxford Street\, Cambridge\, MA\, 02138
CATEGORIES:Programs,Workshop
END:VEVENT
END:VCALENDAR