{"id":199960,"date":"2014-04-02T13:28:34","date_gmt":"2014-04-02T13:28:34","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/events\/new-england-machine-learning-day-2014\/"},"modified":"2025-08-06T12:01:57","modified_gmt":"2025-08-06T19:01:57","slug":"new-england-machine-learning-day-2014","status":"publish","type":"msr-event","link":"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2014\/","title":{"rendered":"New England Machine Learning Day 2014"},"content":{"rendered":"\n\n<p><strong>Venue:<\/strong> <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/lab\/microsoft-research-new-england\/\" target=\"_blank\" rel=\"noopener noreferrer\">Microsoft Research New England<\/a><br \/>\nHorace Mann Conference Room<br \/>\nOne Memorial Drive<br \/>\nCambridge, MA 02142<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>The third annual New England Machine Learning Day will be held May 13, 2014, at Microsoft Research New England, One Memorial Drive, Cambridge, MA 02142. The event will bring together local academics and researchers in machine learning and its applications. See the agenda tab for the list of presentations.<\/p>\n<h2>Related events<\/h2>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2020\/\">NEML 2020<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2019\/\">NEML 2019<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2018\/\">NEML 2018<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2017\/\">NEML 2017<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2016\/\">NEML 2016<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2015\/\">NEML 2015<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2014\/\">NEML 2014<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2013\/\">NEML 2013<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2012\/\">NEML 2012<\/a><\/li>\n<\/ul>\n<p><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>9:50 &#8211; 10:00<br \/>\nOpening remarks<\/p>\n<p>10:00 &#8211; 10:30, Polina Golland, MIT<br \/>\n<em>Joint Modeling of Imaging and Genetic Data<\/em><\/p>\n<p>We propose a unified Bayesian framework for detecting genetic variants associated with a disease while exploiting image-based features as an intermediate phenotype. Traditionally, imaging genetics methods comprise two separate steps. First, image features are selected based on their relevance to the disease phenotype. Second, a set of genetic variants are identified to explain the selected features. In contrast, our method performs these tasks simultaneously to ultimately assign probabilistic measures of relevance to both genetic and imaging markers. We derive an efficient approximate inference algorithm that handles high dimensionality of imaging genetic data. We evaluate the algorithm on synthetic data and show that it outperforms traditional models. We also illustrate the application of the method in a study of Alzheimer&#8217;s disease. Joint work with Kayhan Batmanghelich, Adrian Dalca, Mert Sabuncu.<\/p>\n<p>10:35 &#8211; 11:05, David Cox, Harvard | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-perceptual-annotation-from-biologically-inspired-to-biologically-informed-machine-learning\/\">Video<\/a><br \/>\n<em>&#8220;Perceptual Annotation&#8221;: from Biologically Inspired, to Biologically Informed Machine Learning<\/em><\/p>\n<p>Many machine learning applications, explicitly or implicitly, attempt to mimic natural human abilities in a machine. Indeed, any setting where human-provided labels are used as ground truth \u2013 whether the system aspires to be biologically-inspired or not \u2013 is ultimately driven by the human visual and cognitive system and its ability to provide accurate examplar labels. However, human-provided ground-truth labels are in many ways just the tip of the iceberg of the information that can be extracted from human judgments. I will describe a new approach &#8212; called &#8220;perceptual annotation&#8221; &#8212; in which we use an advanced online psychometric testing platform to acquire new kinds of human annotation data, and we incorporate these data directly into the formulation of a machine learning algorithm. A key intuition for this approach is that while it may be infeasible to dramatically increase the amount of data and high-quality labels available for the training of a given system, measuring the latent exemplar-by-exemplar landscape of difficulty and patterns of human errors can provide important information for regularizing the solution of the system at hand. Finally, I will conclude by exploring how this approach can be extended to incorporate an even greater diversity of different kinds of biological data.<\/p>\n<p>11:10 &#8211; 11:40, Adam Tauman Kalai, Microsoft Research New England | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-better-letters\/\">Video<\/a><br \/>\n<em>Better Letters<\/em><\/p>\n<p>English text today is often machine printed or displayed on screens using the same letters that were carved in stone and handwritten on wax and parchment over two thousand years ago. We consider the problem of developing radically different characters for the same underlying twenty-six letter English alphabet, just as Braille or cursive are alternative representations. We discuss optimizing these letters for multiple criteria using crowdsourcing and machine learning.<\/p>\n<p>11:40 &#8211; 1:45<br \/>\nLunch and posters<\/p>\n<p>1:45 &#8211; 2:15, Michael Littman, Brown University | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-learning-to-act-in-multiagent-sequential-environments\/\">Video<\/a><br \/>\n<em>Learning to Act in Multiagent Sequential Environments<\/em><\/p>\n<p>From routing to online auctions, many decision-making tasks for learning agents are carried out in the presence of other decision makers. I will give a brief overview of results developed in the context of adapting reinforcement-learning algorithms to work effectively in multiagent environments. Of particular interest is the idea that even simple scenarios, such as the well-known Prisoner\u2019s dilemma, require agents to work together, bearing some individual risk, to arrive at mutually beneficial outcomes.<\/p>\n<p>2:20 &#8211; 2:50, Venkatesh Saligrama, Boston University<br \/>\n<em>Signals on Graphs: Efficient Detection & Recovery<\/em><\/p>\n<p>Several problems such as network intrusion, community detection, disease outbreak, and cell signaling can be described in terms of an attributed graph with signals associated with nodes and edges. In these applications presence of intrusion, community, disease outbreak, or signal pathway is characterized by novel observations on some unknown connected subgraph. These problems can be formulated in terms of optimization of suitable objectives on connected sub-graphs, a problem which is generally computationally difficult. We overcome the combinatorics of connectivity algebraically through embedding of connected subgraphs into linear matrix inequalities (LMI). Computationally efficient tests are then realized by optimizing convex objective functions subject to these LMI constraints. We show that our tests are minimax optimal for exponential family of distributions and for graphs satisfying polynomial growth property.<\/p>\n<p>2:50 &#8211; 3:20<br \/>\nCoffee break<\/p>\n<p>3:20 &#8211; 3:50, Carla Brodley, Tufts<br \/>\n<em>Redefining Class Definitions using Constraint-Based Clustering and its Application to Landcover Classification and the AAAI 2014 Keywords<\/em><\/p>\n<p>Two aspects are crucial when constructing any real world supervised classification task: the set of classes whose distinction might be useful for the domain expert, and the set of classifications that can actually be distinguished by the data. Often a set of labels is defined with some initial intuition but these are not the best match for the task. For example, labels have been assigned for land cover classification of the Earth but it has been suspected that these labels are not ideal and some classes may be best split into subclasses whereas others should be merged. We present an approach that formalizes this problem using three ingredients: the existing class labels, the underlying separability in the data, and input from the domain expert specifying an LxL matrix of pairwise probabilistic constraints expressing their beliefs as to whether the L classes should be kept separate, merged, or split. We describe how the problem can be solved by casting it as an instance of constraint-based clustering. We present results demonstrating its application to the task of redefining a class taxonomy for land cover classification of the Earth and redefining the set of high-level keywords for AAAI 2014.<\/p>\n<p>3:55 &#8211; 4:25, Ben Marlin, University of Massachusetts, Amherst<br \/>\n<em>Machine Learning for Clinical and Mobile Health<\/em><\/p>\n<p>The effective analysis of emerging sources of complex clinical and mobile health data represent a key challenge for machine learning. These data often exhibit multiple complicating factors including sparse and irregular sampling, incompleteness, noise, non-stationary temporal dynamics, high levels of between-subjects variability, high volume, high velocity, and significant heterogeneity and multi-modality. In this talk, I will present an overview of some of the machine learning problems my research group is currently working on motivated by ongoing collaborations in both clinical and mobile health. These problems include modeling and prediction of sparse and irregularly sampled physiological time series data from intensive care unit electronic health records, feature extraction and event detection from noisy wearable on-body sensor data, and learning what to sense in the energy and computation constrained mobile device setting.<\/p>\n<p>4:30 &#8211; 5:00, Joshua B Tenenbaum, MIT<br \/>\n<em>Towards More Human-Like Machine Learning<\/em><\/p>\n<p>How can we build a machine that learns to see the world as a human being does? This question has been at the heart of the fields of AI and machine learning since their inceptions. Recently the question has seen renewed interest from researchers taking various &#8220;big data&#8221; approaches, such as training many-layered neural networks to find structure in millions of images, or mining the web to build databases of millions of common-sense facts. I will talk about our recent work taking a different approach, based on trying to reverse-engineer the core cognitive capacities and learning mechanisms of young children and infants. In contrast to conventional big-data ML approaches, children parse their experience using rich causally structured generative models, and learn new models from very little evidence; often just a single example is sufficient to grasp a new concept and generalize in richer ways than machine learning systems can typically do even with hundreds or thousands of examples. I will show how we are beginning to capture these perception and learning abilities in computational terms using techniques based on probabilistic programs and program induction, embedded in a broadly Bayesian framework for inference under uncertainty.<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<table class=\"blue-stripe tWiz tableBorder borderRows borderColumns\" cellpadding=\"7\">\n<tbody>\n<tr class=\"blue-stripeTableHeaderRow\">\n<td class=\"blue-stripeTableHeaderEvenCol\"><strong>Poster Title\u00a0<\/strong><\/td>\n<td class=\"blue-stripeTableHeaderOddCol\"><strong>Presenting Author\/<br \/>\nComplete List of Authors<\/strong><\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Inferring Multilateral Relations from Dynamic Bilateral Interactions<\/td>\n<td class=\"blue-stripeTableOddCol\">Aaron Schein \/ Aaron Schein, Juston Moore, Hanna Wallach<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Sparse Neural Networks and Random-Access Pixel Cameras for Energy Efficient Mobile Gaze Tracking<\/td>\n<td class=\"blue-stripeTableOddCol\">Addison Mayberry \/ Addison Mayberry, Pan Hu, Christopher Salthouse, Benjamin Marlin, Deepak Ganesan<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Relational Dependency Networks for Anomaly Detection<\/td>\n<td class=\"blue-stripeTableOddCol\">Amanda Gentzel \/ Amanda Gentzel, Elisabeth Baseman, Dan Corkill, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Dynamically Generated CRFs for Morphological Analysis of Noisy ECG Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Annamalai Natarajan \/ Annamalai Natarajan, Edward Gaiser, Gustavo Angarita, Robert Malison, Deepak Ganesan, Benjamin Marlin<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Generative and Discriminative Models for Improving Noisy Training Data for Relation Extraction<\/td>\n<td class=\"blue-stripeTableOddCol\">Benjamin Roth \/ Benjamin Roth, Dietrich Klakow<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations<\/td>\n<td class=\"blue-stripeTableOddCol\">Bilal Ahmed \/ Bilal Ahmed, Thomas Thesen, Karen Blackmon, Orrin Devinsky, Ruben Kuzniecky, and Carla E. Brodley<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Boundary algorithm for fast online classification and regression<\/td>\n<td class=\"blue-stripeTableOddCol\">Charles Mathy \/ Charles Mathy, Nate Derbinsky, Jose Bento, Jonathan Rosenthal, Jonathan Yedidia<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Best Response Bayesian Reinforcement Learning for Multiagent Systems with State Uncertainty<\/td>\n<td class=\"blue-stripeTableOddCol\">Chris Amato \/ Frans A. Oliehoek and Christopher Amato<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning Dirichlet Priors for Affordance Aware Planning<\/td>\n<td class=\"blue-stripeTableOddCol\">David Abel and Gabriel Barth-Maron \/ David Abel, Gabriel Barth-Maron, James MacGlashan, Stefanie Tellex<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning with Mixtures of Dependency Networks<\/td>\n<td class=\"blue-stripeTableOddCol\">David Arbour \/ David Arbour, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Employment of Frank-Wolfe algorithm to perform marginal inference in a Gibbs distribution<\/td>\n<td class=\"blue-stripeTableOddCol\">David Belanger<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Restricted Memory Online Variational Bayesian Changepoint Detection<\/td>\n<td class=\"blue-stripeTableOddCol\">Diana Cai \/ Diana Cai, Ryan Adams<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning Modular Structures from Network Data and Node Variables<\/td>\n<td class=\"blue-stripeTableOddCol\">Elham Azizi \/ Elham Azizi, Edoardo M. Airoldi, James E. Galagan<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Dynamic Statistical Models of Collective Social Network Behavior<\/td>\n<td class=\"blue-stripeTableOddCol\">Elisabeth Baseman \/ Elisabeth Baseman, Stephen Judd, Michael Kearns, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Fast Margin-based Cost-sensitive Classification<\/td>\n<td class=\"blue-stripeTableOddCol\">Feng Nan \/ Feng Nan, Joseph Wang, Kirill Trapeznikov, Venkatesh Saligrama<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Multi-Step Stochastic ADMM in High Dimensions: Applications to Sparse Optimization and Noisy Matrix Decomposition<\/td>\n<td class=\"blue-stripeTableOddCol\">Hanie Sedghi<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Augur: a Modeling Language for Data-Parallel Probabilistic Inference<\/td>\n<td class=\"blue-stripeTableOddCol\">Jean-Baptiste Tristan \/ Jean-Baptiste Tristan, Daniel Huang, Joseph Tassarotti, Adam Pocock, Stephen J. Green, Guy L. Steele Jr<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Connected Sub-graph Detection<\/td>\n<td class=\"blue-stripeTableOddCol\">Jing Qian \/ Jing Qian, Venkatesh Saligrama, Yuting Chen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">An information-theoretic analysis of resampling in sequential Monte Carlo<\/td>\n<td class=\"blue-stripeTableOddCol\">Jonathan Huggins \/ Jonathan H. Huggins and Daniel M. Roy<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Text analysis techniques for nominating contact offenders in peer-to-peer file sharing networks<\/td>\n<td class=\"blue-stripeTableOddCol\">Juston Moore \/ Juston Moore, Brian Levine, Marc Liberatore, Hanna Wallach, Janis Wolak<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">A Sound and Complete Algorithm for Learning Causal Models from Relational Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Katerina Marazopoulou \/ Marc Maier, Katerina Marazopoulou, David Arbour, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Time Series Analysis of Mobile Data Usage Reveals Geographic Location<\/td>\n<td class=\"blue-stripeTableOddCol\">Keen Sung \/ Keen Sung, Erik Learned-Miller, Brian Levine, Marc Liberatore<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Evaluating Topic Models through Histogram Analysis<\/td>\n<td class=\"blue-stripeTableOddCol\">Kriste Krstovski \/ Kriste Krstovski, David A. Smith and Michael J. Kurtz<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">A Graphical Model for Entity-based Document Retrieval<\/td>\n<td class=\"blue-stripeTableOddCol\">Laura Dietz \/ Laura Dietz, Jeffrey Dalton, Bruce Croft<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Classifier-Adjusted Density Estimation for Anomaly Detection and One-Class<\/td>\n<td class=\"blue-stripeTableOddCol\">Lisa Friedland \/ Lisa Friedland, Amanda Gentzel, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning of Overcomplete Latent Variable Models: Supervised and Semi-supervised Settings<\/td>\n<td class=\"blue-stripeTableOddCol\">Majid Janzamin<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Tensor Factorization for Large-Scale Relational Learning<\/td>\n<td class=\"blue-stripeTableOddCol\">Maximilian Nickel \/ Maximilian Nickel, Volker Tresp<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Person Re-Identification using Kernel-based Metric Learning Methods<\/td>\n<td class=\"blue-stripeTableOddCol\">Mengran Gou \/ Fei Xiong, Mengran Gou, Octavia Camps, Mario Sznaier<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Regression with No Labeled Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Mohammad Gheshlaghi Azar \/ Mohammad Gheshlaghi Azar and Konrad Kording<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Inferring Helpful Actions<\/td>\n<td class=\"blue-stripeTableOddCol\">Nakul Gopalan \/ Nakul Gopalan, Izaak Baker, Stefanie Tellex<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">DISCOMAX: Distance Correlation Maximization using Graph Laplacians<\/td>\n<td class=\"blue-stripeTableOddCol\">Praneeth Vepakomma \/ Praneeth Vepakomma, Chetan Tonde, Ahmed Elgammal<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Towards Collaborative Filtering Recommender Systems for Tailored Health Communications<\/td>\n<td class=\"blue-stripeTableOddCol\">Roy Adams<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Deterministic Feature Selection for Linear Support Vector Machines<\/td>\n<td class=\"blue-stripeTableOddCol\">Saurabh Paul \/ Saurabh Paul, Malik Magdon-Ismail and Petros Drineas<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">The Value of Temporal Data for Learning Influence Networks<\/td>\n<td class=\"blue-stripeTableOddCol\">Spyros Zoumpoulis \/ Munther Dahleh, John Tsitsiklis, Spyros Zoumpoulis<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Co-Planning via Inverse Reinforcement Learning<\/td>\n<td class=\"blue-stripeTableOddCol\">Stephen Brawner \/ Stephen Brawner, Lee Painton, Stefanie Tellex, Michael Littman<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">A Kernel-Based Framework for Learning with Irregularly Sampled Physiological Time Series<\/td>\n<td class=\"blue-stripeTableOddCol\">Steve Li<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Gradient-based inference for higher-order probabilistic programming languages<\/td>\n<td class=\"blue-stripeTableOddCol\">Tianlin Shi \/ Alexey Radul, Vikash K. Mansinghka<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Sensing-Aware kernel SVMs<\/td>\n<td class=\"blue-stripeTableOddCol\">Weicong Ding \/ Weicong Ding, Prakash Ishwar, Venkatesh Saligrama, W. Clem Karl<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Authorship attribution of unsigned Supreme Court opinions<\/td>\n<td class=\"blue-stripeTableOddCol\">William Li \/ William Li, Pablo Azar, David Larochelle, Phil Hill, James Cox, Robert C. Berwick, Andrew W. Lo<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Modeling and Prediction of Heart-Related Hospitalization Using Electronic Health Records Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Wuyang Dai \/ Wuyang Dai, Theodora Brisimi, Venkatesh Saligrama, Ioannis Ch. Paschalidis<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning dynamic spatiotemporal fields using data from mobile sensors<\/td>\n<td class=\"blue-stripeTableOddCol\">Xiaodong Lan \/ Xiaodong Lan and Mac Schwager<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Handling Physician Subjectivity in the Prediction of Disease Course: An Application to Multiple Sclerosis<\/td>\n<td class=\"blue-stripeTableOddCol\">Yijun Zhao \/ Yijun Zhao, Carla Brodley, Tanuja Chitnis, Brian C. Healy<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">A Convex Moments-based Approach to Subspace Clustering with Priors<\/td>\n<td class=\"blue-stripeTableOddCol\">Yin Wang \/ Yin Wang, Yongfang Cheng, Mario Sznaier, Octavia Camps<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Formal Methods for Learning and Detection of Anomalous Behavior in Cyber-Physcial Systems<\/td>\n<td class=\"blue-stripeTableOddCol\">Zhaodan Kong \/ Zhaodan Kong, Austin Jones, Calin Belta<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>For any questions, please contact <a title=\"\" target=\"_self\">MLday14@microsoft.com<\/a>.<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>Ryan Adams, Computer Science, Harvard<br \/>\nSham Kakade, Microsoft Research New England<br \/>\nLorenzo Rosasco, Universita&#8217; di Genova and MIT<br \/>\nStefanie Tellex, Computer Science, Brown<\/p>\n<div>\n<div class=\"center sw mt16\">\n<div id=\"contentArea\">\n<div id=\"mainZone\" class=\"zone\">\n<div class=\"compositeModule_2_1Zone \">\n<div class=\"zone\">\n<div class=\"conM \">\n<p>The steering committee that selects the organizers of ML Day each year consists of Sham Kakade, Adam Kalai, and Joshua Tenenbaum.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<p><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>The third annual New England Machine Learning Day will be held May 13, 2014, at Microsoft Research New England, One Memorial Drive, Cambridge, MA 02142. The event will bring together local academics and researchers in machine learning and its applications.<\/p>\n","protected":false},"featured_media":0,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_startdate":"2014-05-13","msr_enddate":"2014-05-13","msr_location":"Cambridge, MA, USA","msr_expirationdate":"","msr_event_recording_link":"","msr_event_link":"","msr_event_link_redirect":false,"msr_event_time":"","msr_hide_region":false,"msr_private_event":true,"msr_hide_image_in_river":0,"footnotes":""},"research-area":[13556],"msr-region":[197900],"msr-event-type":[197944],"msr-video-type":[],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[],"msr-impact-theme":[],"class_list":["post-199960","msr-event","type-msr-event","status-publish","hentry","msr-research-area-artificial-intelligence","msr-region-north-america","msr-event-type-hosted-by-microsoft","msr-locale-en_us"],"msr_about":"<!-- wp:msr\/event-details {\"title\":\"New England Machine Learning Day 2014\",\"backgroundColor\":\"grey\"} \/-->\n\n<!-- wp:msr\/content-tabs --><!-- wp:msr\/content-tab {\"title\":\"About\"} --><!-- wp:freeform --><p><strong>Venue:<\/strong> <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/lab\/microsoft-research-new-england\/\" target=\"_blank\" rel=\"noopener noreferrer\">Microsoft Research New England<\/a><br \/>\nHorace Mann Conference Room<br \/>\nOne Memorial Drive<br \/>\nCambridge, MA 02142<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>The third annual New England Machine Learning Day will be held May 13, 2014, at Microsoft Research New England, One Memorial Drive, Cambridge, MA 02142. The event will bring together local academics and researchers in machine learning and its applications. See the agenda tab for the list of presentations.<\/p>\n<h2>Related events<\/h2>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2020\/\">NEML 2020<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2019\/\">NEML 2019<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2018\/\">NEML 2018<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2017\/\">NEML 2017<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2016\/\">NEML 2016<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2015\/\">NEML 2015<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2014\/\">NEML 2014<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2013\/\">NEML 2013<\/a><\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2012\/\">NEML 2012<\/a><\/li>\n<\/ul>\n<p><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- wp:msr\/content-tab {\"title\":\"Agenda\"} --><!-- wp:freeform --><p>9:50 &#8211; 10:00<br \/>\nOpening remarks<\/p>\n<p>10:00 &#8211; 10:30, Polina Golland, MIT<br \/>\n<em>Joint Modeling of Imaging and Genetic Data<\/em><\/p>\n<p>We propose a unified Bayesian framework for detecting genetic variants associated with a disease while exploiting image-based features as an intermediate phenotype. Traditionally, imaging genetics methods comprise two separate steps. First, image features are selected based on their relevance to the disease phenotype. Second, a set of genetic variants are identified to explain the selected features. In contrast, our method performs these tasks simultaneously to ultimately assign probabilistic measures of relevance to both genetic and imaging markers. We derive an efficient approximate inference algorithm that handles high dimensionality of imaging genetic data. We evaluate the algorithm on synthetic data and show that it outperforms traditional models. We also illustrate the application of the method in a study of Alzheimer&#8217;s disease. Joint work with Kayhan Batmanghelich, Adrian Dalca, Mert Sabuncu.<\/p>\n<p>10:35 &#8211; 11:05, David Cox, Harvard | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-perceptual-annotation-from-biologically-inspired-to-biologically-informed-machine-learning\/\">Video<\/a><br \/>\n<em>&#8220;Perceptual Annotation&#8221;: from Biologically Inspired, to Biologically Informed Machine Learning<\/em><\/p>\n<p>Many machine learning applications, explicitly or implicitly, attempt to mimic natural human abilities in a machine. Indeed, any setting where human-provided labels are used as ground truth \u2013 whether the system aspires to be biologically-inspired or not \u2013 is ultimately driven by the human visual and cognitive system and its ability to provide accurate examplar labels. However, human-provided ground-truth labels are in many ways just the tip of the iceberg of the information that can be extracted from human judgments. I will describe a new approach &#8212; called &#8220;perceptual annotation&#8221; &#8212; in which we use an advanced online psychometric testing platform to acquire new kinds of human annotation data, and we incorporate these data directly into the formulation of a machine learning algorithm. A key intuition for this approach is that while it may be infeasible to dramatically increase the amount of data and high-quality labels available for the training of a given system, measuring the latent exemplar-by-exemplar landscape of difficulty and patterns of human errors can provide important information for regularizing the solution of the system at hand. Finally, I will conclude by exploring how this approach can be extended to incorporate an even greater diversity of different kinds of biological data.<\/p>\n<p>11:10 &#8211; 11:40, Adam Tauman Kalai, Microsoft Research New England | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-better-letters\/\">Video<\/a><br \/>\n<em>Better Letters<\/em><\/p>\n<p>English text today is often machine printed or displayed on screens using the same letters that were carved in stone and handwritten on wax and parchment over two thousand years ago. We consider the problem of developing radically different characters for the same underlying twenty-six letter English alphabet, just as Braille or cursive are alternative representations. We discuss optimizing these letters for multiple criteria using crowdsourcing and machine learning.<\/p>\n<p>11:40 &#8211; 1:45<br \/>\nLunch and posters<\/p>\n<p>1:45 &#8211; 2:15, Michael Littman, Brown University | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-learning-to-act-in-multiagent-sequential-environments\/\">Video<\/a><br \/>\n<em>Learning to Act in Multiagent Sequential Environments<\/em><\/p>\n<p>From routing to online auctions, many decision-making tasks for learning agents are carried out in the presence of other decision makers. I will give a brief overview of results developed in the context of adapting reinforcement-learning algorithms to work effectively in multiagent environments. Of particular interest is the idea that even simple scenarios, such as the well-known Prisoner\u2019s dilemma, require agents to work together, bearing some individual risk, to arrive at mutually beneficial outcomes.<\/p>\n<p>2:20 &#8211; 2:50, Venkatesh Saligrama, Boston University<br \/>\n<em>Signals on Graphs: Efficient Detection &amp; Recovery<\/em><\/p>\n<p>Several problems such as network intrusion, community detection, disease outbreak, and cell signaling can be described in terms of an attributed graph with signals associated with nodes and edges. In these applications presence of intrusion, community, disease outbreak, or signal pathway is characterized by novel observations on some unknown connected subgraph. These problems can be formulated in terms of optimization of suitable objectives on connected sub-graphs, a problem which is generally computationally difficult. We overcome the combinatorics of connectivity algebraically through embedding of connected subgraphs into linear matrix inequalities (LMI). Computationally efficient tests are then realized by optimizing convex objective functions subject to these LMI constraints. We show that our tests are minimax optimal for exponential family of distributions and for graphs satisfying polynomial growth property.<\/p>\n<p>2:50 &#8211; 3:20<br \/>\nCoffee break<\/p>\n<p>3:20 &#8211; 3:50, Carla Brodley, Tufts<br \/>\n<em>Redefining Class Definitions using Constraint-Based Clustering and its Application to Landcover Classification and the AAAI 2014 Keywords<\/em><\/p>\n<p>Two aspects are crucial when constructing any real world supervised classification task: the set of classes whose distinction might be useful for the domain expert, and the set of classifications that can actually be distinguished by the data. Often a set of labels is defined with some initial intuition but these are not the best match for the task. For example, labels have been assigned for land cover classification of the Earth but it has been suspected that these labels are not ideal and some classes may be best split into subclasses whereas others should be merged. We present an approach that formalizes this problem using three ingredients: the existing class labels, the underlying separability in the data, and input from the domain expert specifying an LxL matrix of pairwise probabilistic constraints expressing their beliefs as to whether the L classes should be kept separate, merged, or split. We describe how the problem can be solved by casting it as an instance of constraint-based clustering. We present results demonstrating its application to the task of redefining a class taxonomy for land cover classification of the Earth and redefining the set of high-level keywords for AAAI 2014.<\/p>\n<p>3:55 &#8211; 4:25, Ben Marlin, University of Massachusetts, Amherst<br \/>\n<em>Machine Learning for Clinical and Mobile Health<\/em><\/p>\n<p>The effective analysis of emerging sources of complex clinical and mobile health data represent a key challenge for machine learning. These data often exhibit multiple complicating factors including sparse and irregular sampling, incompleteness, noise, non-stationary temporal dynamics, high levels of between-subjects variability, high volume, high velocity, and significant heterogeneity and multi-modality. In this talk, I will present an overview of some of the machine learning problems my research group is currently working on motivated by ongoing collaborations in both clinical and mobile health. These problems include modeling and prediction of sparse and irregularly sampled physiological time series data from intensive care unit electronic health records, feature extraction and event detection from noisy wearable on-body sensor data, and learning what to sense in the energy and computation constrained mobile device setting.<\/p>\n<p>4:30 &#8211; 5:00, Joshua B Tenenbaum, MIT<br \/>\n<em>Towards More Human-Like Machine Learning<\/em><\/p>\n<p>How can we build a machine that learns to see the world as a human being does? This question has been at the heart of the fields of AI and machine learning since their inceptions. Recently the question has seen renewed interest from researchers taking various &#8220;big data&#8221; approaches, such as training many-layered neural networks to find structure in millions of images, or mining the web to build databases of millions of common-sense facts. I will talk about our recent work taking a different approach, based on trying to reverse-engineer the core cognitive capacities and learning mechanisms of young children and infants. In contrast to conventional big-data ML approaches, children parse their experience using rich causally structured generative models, and learn new models from very little evidence; often just a single example is sufficient to grasp a new concept and generalize in richer ways than machine learning systems can typically do even with hundreds or thousands of examples. I will show how we are beginning to capture these perception and learning abilities in computational terms using techniques based on probabilistic programs and program induction, embedded in a broadly Bayesian framework for inference under uncertainty.<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- wp:msr\/content-tab {\"title\":\"Posters\"} --><!-- wp:freeform --><table class=\"blue-stripe tWiz tableBorder borderRows borderColumns\" cellpadding=\"7\">\n<tbody>\n<tr class=\"blue-stripeTableHeaderRow\">\n<td class=\"blue-stripeTableHeaderEvenCol\"><strong>Poster Title\u00a0<\/strong><\/td>\n<td class=\"blue-stripeTableHeaderOddCol\"><strong>Presenting Author\/<br \/>\nComplete List of Authors<\/strong><\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Inferring Multilateral Relations from Dynamic Bilateral Interactions<\/td>\n<td class=\"blue-stripeTableOddCol\">Aaron Schein \/ Aaron Schein, Juston Moore, Hanna Wallach<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Sparse Neural Networks and Random-Access Pixel Cameras for Energy Efficient Mobile Gaze Tracking<\/td>\n<td class=\"blue-stripeTableOddCol\">Addison Mayberry \/ Addison Mayberry, Pan Hu, Christopher Salthouse, Benjamin Marlin, Deepak Ganesan<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Relational Dependency Networks for Anomaly Detection<\/td>\n<td class=\"blue-stripeTableOddCol\">Amanda Gentzel \/ Amanda Gentzel, Elisabeth Baseman, Dan Corkill, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Dynamically Generated CRFs for Morphological Analysis of Noisy ECG Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Annamalai Natarajan \/ Annamalai Natarajan, Edward Gaiser, Gustavo Angarita, Robert Malison, Deepak Ganesan, Benjamin Marlin<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Generative and Discriminative Models for Improving Noisy Training Data for Relation Extraction<\/td>\n<td class=\"blue-stripeTableOddCol\">Benjamin Roth \/ Benjamin Roth, Dietrich Klakow<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations<\/td>\n<td class=\"blue-stripeTableOddCol\">Bilal Ahmed \/ Bilal Ahmed, Thomas Thesen, Karen Blackmon, Orrin Devinsky, Ruben Kuzniecky, and Carla E. Brodley<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Boundary algorithm for fast online classification and regression<\/td>\n<td class=\"blue-stripeTableOddCol\">Charles Mathy \/ Charles Mathy, Nate Derbinsky, Jose Bento, Jonathan Rosenthal, Jonathan Yedidia<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Best Response Bayesian Reinforcement Learning for Multiagent Systems with State Uncertainty<\/td>\n<td class=\"blue-stripeTableOddCol\">Chris Amato \/ Frans A. Oliehoek and Christopher Amato<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning Dirichlet Priors for Affordance Aware Planning<\/td>\n<td class=\"blue-stripeTableOddCol\">David Abel and Gabriel Barth-Maron \/ David Abel, Gabriel Barth-Maron, James MacGlashan, Stefanie Tellex<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning with Mixtures of Dependency Networks<\/td>\n<td class=\"blue-stripeTableOddCol\">David Arbour \/ David Arbour, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Employment of Frank-Wolfe algorithm to perform marginal inference in a Gibbs distribution<\/td>\n<td class=\"blue-stripeTableOddCol\">David Belanger<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Restricted Memory Online Variational Bayesian Changepoint Detection<\/td>\n<td class=\"blue-stripeTableOddCol\">Diana Cai \/ Diana Cai, Ryan Adams<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning Modular Structures from Network Data and Node Variables<\/td>\n<td class=\"blue-stripeTableOddCol\">Elham Azizi \/ Elham Azizi, Edoardo M. Airoldi, James E. Galagan<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Dynamic Statistical Models of Collective Social Network Behavior<\/td>\n<td class=\"blue-stripeTableOddCol\">Elisabeth Baseman \/ Elisabeth Baseman, Stephen Judd, Michael Kearns, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Fast Margin-based Cost-sensitive Classification<\/td>\n<td class=\"blue-stripeTableOddCol\">Feng Nan \/ Feng Nan, Joseph Wang, Kirill Trapeznikov, Venkatesh Saligrama<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Multi-Step Stochastic ADMM in High Dimensions: Applications to Sparse Optimization and Noisy Matrix Decomposition<\/td>\n<td class=\"blue-stripeTableOddCol\">Hanie Sedghi<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Augur: a Modeling Language for Data-Parallel Probabilistic Inference<\/td>\n<td class=\"blue-stripeTableOddCol\">Jean-Baptiste Tristan \/ Jean-Baptiste Tristan, Daniel Huang, Joseph Tassarotti, Adam Pocock, Stephen J. Green, Guy L. Steele Jr<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Connected Sub-graph Detection<\/td>\n<td class=\"blue-stripeTableOddCol\">Jing Qian \/ Jing Qian, Venkatesh Saligrama, Yuting Chen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">An information-theoretic analysis of resampling in sequential Monte Carlo<\/td>\n<td class=\"blue-stripeTableOddCol\">Jonathan Huggins \/ Jonathan H. Huggins and Daniel M. Roy<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Text analysis techniques for nominating contact offenders in peer-to-peer file sharing networks<\/td>\n<td class=\"blue-stripeTableOddCol\">Juston Moore \/ Juston Moore, Brian Levine, Marc Liberatore, Hanna Wallach, Janis Wolak<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">A Sound and Complete Algorithm for Learning Causal Models from Relational Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Katerina Marazopoulou \/ Marc Maier, Katerina Marazopoulou, David Arbour, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Time Series Analysis of Mobile Data Usage Reveals Geographic Location<\/td>\n<td class=\"blue-stripeTableOddCol\">Keen Sung \/ Keen Sung, Erik Learned-Miller, Brian Levine, Marc Liberatore<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Evaluating Topic Models through Histogram Analysis<\/td>\n<td class=\"blue-stripeTableOddCol\">Kriste Krstovski \/ Kriste Krstovski, David A. Smith and Michael J. Kurtz<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">A Graphical Model for Entity-based Document Retrieval<\/td>\n<td class=\"blue-stripeTableOddCol\">Laura Dietz \/ Laura Dietz, Jeffrey Dalton, Bruce Croft<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Classifier-Adjusted Density Estimation for Anomaly Detection and One-Class<\/td>\n<td class=\"blue-stripeTableOddCol\">Lisa Friedland \/ Lisa Friedland, Amanda Gentzel, David Jensen<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning of Overcomplete Latent Variable Models: Supervised and Semi-supervised Settings<\/td>\n<td class=\"blue-stripeTableOddCol\">Majid Janzamin<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Tensor Factorization for Large-Scale Relational Learning<\/td>\n<td class=\"blue-stripeTableOddCol\">Maximilian Nickel \/ Maximilian Nickel, Volker Tresp<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Person Re-Identification using Kernel-based Metric Learning Methods<\/td>\n<td class=\"blue-stripeTableOddCol\">Mengran Gou \/ Fei Xiong, Mengran Gou, Octavia Camps, Mario Sznaier<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Regression with No Labeled Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Mohammad Gheshlaghi Azar \/ Mohammad Gheshlaghi Azar and Konrad Kording<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Inferring Helpful Actions<\/td>\n<td class=\"blue-stripeTableOddCol\">Nakul Gopalan \/ Nakul Gopalan, Izaak Baker, Stefanie Tellex<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">DISCOMAX: Distance Correlation Maximization using Graph Laplacians<\/td>\n<td class=\"blue-stripeTableOddCol\">Praneeth Vepakomma \/ Praneeth Vepakomma, Chetan Tonde, Ahmed Elgammal<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Towards Collaborative Filtering Recommender Systems for Tailored Health Communications<\/td>\n<td class=\"blue-stripeTableOddCol\">Roy Adams<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Deterministic Feature Selection for Linear Support Vector Machines<\/td>\n<td class=\"blue-stripeTableOddCol\">Saurabh Paul \/ Saurabh Paul, Malik Magdon-Ismail and Petros Drineas<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">The Value of Temporal Data for Learning Influence Networks<\/td>\n<td class=\"blue-stripeTableOddCol\">Spyros Zoumpoulis \/ Munther Dahleh, John Tsitsiklis, Spyros Zoumpoulis<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Co-Planning via Inverse Reinforcement Learning<\/td>\n<td class=\"blue-stripeTableOddCol\">Stephen Brawner \/ Stephen Brawner, Lee Painton, Stefanie Tellex, Michael Littman<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">A Kernel-Based Framework for Learning with Irregularly Sampled Physiological Time Series<\/td>\n<td class=\"blue-stripeTableOddCol\">Steve Li<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Gradient-based inference for higher-order probabilistic programming languages<\/td>\n<td class=\"blue-stripeTableOddCol\">Tianlin Shi \/ Alexey Radul, Vikash K. Mansinghka<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Sensing-Aware kernel SVMs<\/td>\n<td class=\"blue-stripeTableOddCol\">Weicong Ding \/ Weicong Ding, Prakash Ishwar, Venkatesh Saligrama, W. Clem Karl<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Authorship attribution of unsigned Supreme Court opinions<\/td>\n<td class=\"blue-stripeTableOddCol\">William Li \/ William Li, Pablo Azar, David Larochelle, Phil Hill, James Cox, Robert C. Berwick, Andrew W. Lo<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Modeling and Prediction of Heart-Related Hospitalization Using Electronic Health Records Data<\/td>\n<td class=\"blue-stripeTableOddCol\">Wuyang Dai \/ Wuyang Dai, Theodora Brisimi, Venkatesh Saligrama, Ioannis Ch. Paschalidis<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">Learning dynamic spatiotemporal fields using data from mobile sensors<\/td>\n<td class=\"blue-stripeTableOddCol\">Xiaodong Lan \/ Xiaodong Lan and Mac Schwager<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Handling Physician Subjectivity in the Prediction of Disease Course: An Application to Multiple Sclerosis<\/td>\n<td class=\"blue-stripeTableOddCol\">Yijun Zhao \/ Yijun Zhao, Carla Brodley, Tanuja Chitnis, Brian C. Healy<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableOddRow\">\n<td class=\"blue-stripeTableEvenCol\">A Convex Moments-based Approach to Subspace Clustering with Priors<\/td>\n<td class=\"blue-stripeTableOddCol\">Yin Wang \/ Yin Wang, Yongfang Cheng, Mario Sznaier, Octavia Camps<\/td>\n<\/tr>\n<tr class=\"blue-stripeTableEvenRow\">\n<td class=\"blue-stripeTableEvenCol\">Formal Methods for Learning and Detection of Anomalous Behavior in Cyber-Physcial Systems<\/td>\n<td class=\"blue-stripeTableOddCol\">Zhaodan Kong \/ Zhaodan Kong, Austin Jones, Calin Belta<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>For any questions, please contact <a title=\"\" target=\"_self\">MLday14@microsoft.com<\/a>.<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- wp:msr\/content-tab {\"title\":\"Committee\"} --><!-- wp:freeform --><p>Ryan Adams, Computer Science, Harvard<br \/>\nSham Kakade, Microsoft Research New England<br \/>\nLorenzo Rosasco, Universita&#8217; di Genova and MIT<br \/>\nStefanie Tellex, Computer Science, Brown<\/p>\n<div>\n<div class=\"center sw mt16\">\n<div id=\"contentArea\">\n<div id=\"mainZone\" class=\"zone\">\n<div class=\"compositeModule_2_1Zone \">\n<div class=\"zone\">\n<div class=\"conM \">\n<p>The steering committee that selects the organizers of ML Day each year consists of Sham Kakade, Adam Kalai, and Joshua Tenenbaum.<\/p>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<\/div>\n<p><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<!-- \/wp:freeform --><!-- \/wp:msr\/content-tab --><!-- \/wp:msr\/content-tabs -->","tab-content":[{"id":0,"name":"About","content":"The third annual New England Machine Learning Day will be held May 13, 2014, at Microsoft Research New England, One Memorial Drive, Cambridge, MA 02142. The event will bring together local academics and researchers in machine learning and its applications. See the agenda tab for the list of presentations.\r\n<h2>Related events<\/h2>\r\n<ul>\r\n        <li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2020\/\">NEML 2020<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2019\/\">NEML 2019<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2018\/\">NEML 2018<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2017\/\">NEML 2017<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2016\/\">NEML 2016<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2015\/\">NEML 2015<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2014\/\">NEML 2014<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2013\/\">NEML 2013<\/a><\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/event\/new-england-machine-learning-day-2012\/\">NEML 2012<\/a><\/li>\r\n<\/ul>"},{"id":1,"name":"Agenda","content":"9:50 - 10:00\r\nOpening remarks\r\n\r\n10:00 - 10:30, Polina Golland, MIT\r\n<em>Joint Modeling of Imaging and Genetic Data<\/em>\r\n\r\nWe propose a unified Bayesian framework for detecting genetic variants associated with a disease while exploiting image-based features as an intermediate phenotype. Traditionally, imaging genetics methods comprise two separate steps. First, image features are selected based on their relevance to the disease phenotype. Second, a set of genetic variants are identified to explain the selected features. In contrast, our method performs these tasks simultaneously to ultimately assign probabilistic measures of relevance to both genetic and imaging markers. We derive an efficient approximate inference algorithm that handles high dimensionality of imaging genetic data. We evaluate the algorithm on synthetic data and show that it outperforms traditional models. We also illustrate the application of the method in a study of Alzheimer's disease. Joint work with Kayhan Batmanghelich, Adrian Dalca, Mert Sabuncu.\r\n\r\n10:35 - 11:05, David Cox, Harvard | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-perceptual-annotation-from-biologically-inspired-to-biologically-informed-machine-learning\/\">Video<\/a>\r\n<em>\"Perceptual Annotation\": from Biologically Inspired, to Biologically Informed Machine Learning<\/em>\r\n\r\nMany machine learning applications, explicitly or implicitly, attempt to mimic natural human abilities in a machine. Indeed, any setting where human-provided labels are used as ground truth \u2013 whether the system aspires to be biologically-inspired or not \u2013 is ultimately driven by the human visual and cognitive system and its ability to provide accurate examplar labels. However, human-provided ground-truth labels are in many ways just the tip of the iceberg of the information that can be extracted from human judgments. I will describe a new approach -- called \"perceptual annotation\" -- in which we use an advanced online psychometric testing platform to acquire new kinds of human annotation data, and we incorporate these data directly into the formulation of a machine learning algorithm. A key intuition for this approach is that while it may be infeasible to dramatically increase the amount of data and high-quality labels available for the training of a given system, measuring the latent exemplar-by-exemplar landscape of difficulty and patterns of human errors can provide important information for regularizing the solution of the system at hand. Finally, I will conclude by exploring how this approach can be extended to incorporate an even greater diversity of different kinds of biological data.\r\n\r\n11:10 - 11:40, Adam Tauman Kalai, Microsoft Research New England | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-better-letters\/\">Video<\/a>\r\n<em>Better Letters<\/em>\r\n\r\nEnglish text today is often machine printed or displayed on screens using the same letters that were carved in stone and handwritten on wax and parchment over two thousand years ago. We consider the problem of developing radically different characters for the same underlying twenty-six letter English alphabet, just as Braille or cursive are alternative representations. We discuss optimizing these letters for multiple criteria using crowdsourcing and machine learning.\r\n\r\n11:40 - 1:45\r\nLunch and posters\r\n\r\n1:45 - 2:15, Michael Littman, Brown University | <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/video\/ml-day-2014-learning-to-act-in-multiagent-sequential-environments\/\">Video<\/a>\r\n<em>Learning to Act in Multiagent Sequential Environments<\/em>\r\n\r\nFrom routing to online auctions, many decision-making tasks for learning agents are carried out in the presence of other decision makers. I will give a brief overview of results developed in the context of adapting reinforcement-learning algorithms to work effectively in multiagent environments. Of particular interest is the idea that even simple scenarios, such as the well-known Prisoner\u2019s dilemma, require agents to work together, bearing some individual risk, to arrive at mutually beneficial outcomes.\r\n\r\n2:20 - 2:50, Venkatesh Saligrama, Boston University\r\n<em>Signals on Graphs: Efficient Detection &amp; Recovery<\/em>\r\n\r\nSeveral problems such as network intrusion, community detection, disease outbreak, and cell signaling can be described in terms of an attributed graph with signals associated with nodes and edges. In these applications presence of intrusion, community, disease outbreak, or signal pathway is characterized by novel observations on some unknown connected subgraph. These problems can be formulated in terms of optimization of suitable objectives on connected sub-graphs, a problem which is generally computationally difficult. We overcome the combinatorics of connectivity algebraically through embedding of connected subgraphs into linear matrix inequalities (LMI). Computationally efficient tests are then realized by optimizing convex objective functions subject to these LMI constraints. We show that our tests are minimax optimal for exponential family of distributions and for graphs satisfying polynomial growth property.\r\n\r\n2:50 - 3:20\r\nCoffee break\r\n\r\n3:20 - 3:50, Carla Brodley, Tufts\r\n<em>Redefining Class Definitions using Constraint-Based Clustering and its Application to Landcover Classification and the AAAI 2014 Keywords<\/em>\r\n\r\nTwo aspects are crucial when constructing any real world supervised classification task: the set of classes whose distinction might be useful for the domain expert, and the set of classifications that can actually be distinguished by the data. Often a set of labels is defined with some initial intuition but these are not the best match for the task. For example, labels have been assigned for land cover classification of the Earth but it has been suspected that these labels are not ideal and some classes may be best split into subclasses whereas others should be merged. We present an approach that formalizes this problem using three ingredients: the existing class labels, the underlying separability in the data, and input from the domain expert specifying an LxL matrix of pairwise probabilistic constraints expressing their beliefs as to whether the L classes should be kept separate, merged, or split. We describe how the problem can be solved by casting it as an instance of constraint-based clustering. We present results demonstrating its application to the task of redefining a class taxonomy for land cover classification of the Earth and redefining the set of high-level keywords for AAAI 2014.\r\n\r\n3:55 - 4:25, Ben Marlin, University of Massachusetts, Amherst\r\n<em>Machine Learning for Clinical and Mobile Health<\/em>\r\n\r\nThe effective analysis of emerging sources of complex clinical and mobile health data represent a key challenge for machine learning. These data often exhibit multiple complicating factors including sparse and irregular sampling, incompleteness, noise, non-stationary temporal dynamics, high levels of between-subjects variability, high volume, high velocity, and significant heterogeneity and multi-modality. In this talk, I will present an overview of some of the machine learning problems my research group is currently working on motivated by ongoing collaborations in both clinical and mobile health. These problems include modeling and prediction of sparse and irregularly sampled physiological time series data from intensive care unit electronic health records, feature extraction and event detection from noisy wearable on-body sensor data, and learning what to sense in the energy and computation constrained mobile device setting.\r\n\r\n4:30 - 5:00, Joshua B Tenenbaum, MIT\r\n<em>Towards More Human-Like Machine Learning<\/em>\r\n\r\nHow can we build a machine that learns to see the world as a human being does? This question has been at the heart of the fields of AI and machine learning since their inceptions. Recently the question has seen renewed interest from researchers taking various \"big data\" approaches, such as training many-layered neural networks to find structure in millions of images, or mining the web to build databases of millions of common-sense facts. I will talk about our recent work taking a different approach, based on trying to reverse-engineer the core cognitive capacities and learning mechanisms of young children and infants. In contrast to conventional big-data ML approaches, children parse their experience using rich causally structured generative models, and learn new models from very little evidence; often just a single example is sufficient to grasp a new concept and generalize in richer ways than machine learning systems can typically do even with hundreds or thousands of examples. I will show how we are beginning to capture these perception and learning abilities in computational terms using techniques based on probabilistic programs and program induction, embedded in a broadly Bayesian framework for inference under uncertainty."},{"id":2,"name":"Posters","content":"<table class=\"blue-stripe tWiz tableBorder borderRows borderColumns\" cellpadding=\"7\">\r\n<tbody>\r\n<tr class=\"blue-stripeTableHeaderRow\">\r\n<td class=\"blue-stripeTableHeaderEvenCol\"><strong>Poster Title\u00a0<\/strong><\/td>\r\n<td class=\"blue-stripeTableHeaderOddCol\"><strong>Presenting Author\/\r\nComplete List of Authors<\/strong><\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Inferring Multilateral Relations from Dynamic Bilateral Interactions<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Aaron Schein \/ Aaron Schein, Juston Moore, Hanna Wallach<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Sparse Neural Networks and Random-Access Pixel Cameras for Energy Efficient Mobile Gaze Tracking<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Addison Mayberry \/ Addison Mayberry, Pan Hu, Christopher Salthouse, Benjamin Marlin, Deepak Ganesan<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Relational Dependency Networks for Anomaly Detection<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Amanda Gentzel \/ Amanda Gentzel, Elisabeth Baseman, Dan Corkill, David Jensen<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Dynamically Generated CRFs for Morphological Analysis of Noisy ECG Data<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Annamalai Natarajan \/ Annamalai Natarajan, Edward Gaiser, Gustavo Angarita, Robert Malison, Deepak Ganesan, Benjamin Marlin<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Generative and Discriminative Models for Improving Noisy Training Data for Relation Extraction<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Benjamin Roth \/ Benjamin Roth, Dietrich Klakow<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Hierarchical Conditional Random Fields for Outlier Detection: An Application to Detecting Epileptogenic Cortical Malformations<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Bilal Ahmed \/ Bilal Ahmed, Thomas Thesen, Karen Blackmon, Orrin Devinsky, Ruben Kuzniecky, and Carla E. Brodley<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Boundary algorithm for fast online classification and regression<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Charles Mathy \/ Charles Mathy, Nate Derbinsky, Jose Bento, Jonathan Rosenthal, Jonathan Yedidia<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Best Response Bayesian Reinforcement Learning for Multiagent Systems with State Uncertainty<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Chris Amato \/ Frans A. Oliehoek and Christopher Amato<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Learning Dirichlet Priors for Affordance Aware Planning<\/td>\r\n<td class=\"blue-stripeTableOddCol\">David Abel and Gabriel Barth-Maron \/ David Abel, Gabriel Barth-Maron, James MacGlashan, Stefanie Tellex<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Learning with Mixtures of Dependency Networks<\/td>\r\n<td class=\"blue-stripeTableOddCol\">David Arbour \/ David Arbour, David Jensen<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Employment of Frank-Wolfe algorithm to perform marginal inference in a Gibbs distribution<\/td>\r\n<td class=\"blue-stripeTableOddCol\">David Belanger<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Restricted Memory Online Variational Bayesian Changepoint Detection<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Diana Cai \/ Diana Cai, Ryan Adams<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Learning Modular Structures from Network Data and Node Variables<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Elham Azizi \/ Elham Azizi, Edoardo M. Airoldi, James E. Galagan<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Dynamic Statistical Models of Collective Social Network Behavior<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Elisabeth Baseman \/ Elisabeth Baseman, Stephen Judd, Michael Kearns, David Jensen<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Fast Margin-based Cost-sensitive Classification<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Feng Nan \/ Feng Nan, Joseph Wang, Kirill Trapeznikov, Venkatesh Saligrama<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Multi-Step Stochastic ADMM in High Dimensions: Applications to Sparse Optimization and Noisy Matrix Decomposition<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Hanie Sedghi<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Augur: a Modeling Language for Data-Parallel Probabilistic Inference<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Jean-Baptiste Tristan \/ Jean-Baptiste Tristan, Daniel Huang, Joseph Tassarotti, Adam Pocock, Stephen J. Green, Guy L. Steele Jr<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Connected Sub-graph Detection<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Jing Qian \/ Jing Qian, Venkatesh Saligrama, Yuting Chen<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">An information-theoretic analysis of resampling in sequential Monte Carlo<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Jonathan Huggins \/ Jonathan H. Huggins and Daniel M. Roy<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Text analysis techniques for nominating contact offenders in peer-to-peer file sharing networks<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Juston Moore \/ Juston Moore, Brian Levine, Marc Liberatore, Hanna Wallach, Janis Wolak<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">A Sound and Complete Algorithm for Learning Causal Models from Relational Data<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Katerina Marazopoulou \/ Marc Maier, Katerina Marazopoulou, David Arbour, David Jensen<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Time Series Analysis of Mobile Data Usage Reveals Geographic Location<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Keen Sung \/ Keen Sung, Erik Learned-Miller, Brian Levine, Marc Liberatore<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Evaluating Topic Models through Histogram Analysis<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Kriste Krstovski \/ Kriste Krstovski, David A. Smith and Michael J. Kurtz<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">A Graphical Model for Entity-based Document Retrieval<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Laura Dietz \/ Laura Dietz, Jeffrey Dalton, Bruce Croft<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Classifier-Adjusted Density Estimation for Anomaly Detection and One-Class<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Lisa Friedland \/ Lisa Friedland, Amanda Gentzel, David Jensen<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Learning of Overcomplete Latent Variable Models: Supervised and Semi-supervised Settings<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Majid Janzamin<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Tensor Factorization for Large-Scale Relational Learning<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Maximilian Nickel \/ Maximilian Nickel, Volker Tresp<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Person Re-Identification using Kernel-based Metric Learning Methods<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Mengran Gou \/ Fei Xiong, Mengran Gou, Octavia Camps, Mario Sznaier<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Regression with No Labeled Data<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Mohammad Gheshlaghi Azar \/ Mohammad Gheshlaghi Azar and Konrad Kording<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Inferring Helpful Actions<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Nakul Gopalan \/ Nakul Gopalan, Izaak Baker, Stefanie Tellex<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">DISCOMAX: Distance Correlation Maximization using Graph Laplacians<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Praneeth Vepakomma \/ Praneeth Vepakomma, Chetan Tonde, Ahmed Elgammal<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Towards Collaborative Filtering Recommender Systems for Tailored Health Communications<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Roy Adams<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Deterministic Feature Selection for Linear Support Vector Machines<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Saurabh Paul \/ Saurabh Paul, Malik Magdon-Ismail and Petros Drineas<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">The Value of Temporal Data for Learning Influence Networks<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Spyros Zoumpoulis \/ Munther Dahleh, John Tsitsiklis, Spyros Zoumpoulis<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Co-Planning via Inverse Reinforcement Learning<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Stephen Brawner \/ Stephen Brawner, Lee Painton, Stefanie Tellex, Michael Littman<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">A Kernel-Based Framework for Learning with Irregularly Sampled Physiological Time Series<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Steve Li<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Gradient-based inference for higher-order probabilistic programming languages<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Tianlin Shi \/ Alexey Radul, Vikash K. Mansinghka<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Sensing-Aware kernel SVMs<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Weicong Ding \/ Weicong Ding, Prakash Ishwar, Venkatesh Saligrama, W. Clem Karl<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Authorship attribution of unsigned Supreme Court opinions<\/td>\r\n<td class=\"blue-stripeTableOddCol\">William Li \/ William Li, Pablo Azar, David Larochelle, Phil Hill, James Cox, Robert C. Berwick, Andrew W. Lo<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Modeling and Prediction of Heart-Related Hospitalization Using Electronic Health Records Data<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Wuyang Dai \/ Wuyang Dai, Theodora Brisimi, Venkatesh Saligrama, Ioannis Ch. Paschalidis<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Learning dynamic spatiotemporal fields using data from mobile sensors<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Xiaodong Lan \/ Xiaodong Lan and Mac Schwager<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Handling Physician Subjectivity in the Prediction of Disease Course: An Application to Multiple Sclerosis<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Yijun Zhao \/ Yijun Zhao, Carla Brodley, Tanuja Chitnis, Brian C. Healy<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableOddRow\">\r\n<td class=\"blue-stripeTableEvenCol\">A Convex Moments-based Approach to Subspace Clustering with Priors<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Yin Wang \/ Yin Wang, Yongfang Cheng, Mario Sznaier, Octavia Camps<\/td>\r\n<\/tr>\r\n<tr class=\"blue-stripeTableEvenRow\">\r\n<td class=\"blue-stripeTableEvenCol\">Formal Methods for Learning and Detection of Anomalous Behavior in Cyber-Physcial Systems<\/td>\r\n<td class=\"blue-stripeTableOddCol\">Zhaodan Kong \/ Zhaodan Kong, Austin Jones, Calin Belta<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table>\r\nFor any questions, please contact <a title=\"\" target=\"_self\">MLday14@microsoft.com<\/a>."},{"id":3,"name":"Committee","content":"Ryan Adams, Computer Science, Harvard\r\nSham Kakade, Microsoft Research New England\r\nLorenzo Rosasco, Universita' di Genova and MIT\r\nStefanie Tellex, Computer Science, Brown\r\n<div>\r\n<div class=\"center sw mt16\">\r\n<div id=\"contentArea\">\r\n<div id=\"mainZone\" class=\"zone\">\r\n<div class=\"compositeModule_2_1Zone \">\r\n<div class=\"zone\">\r\n<div class=\"conM \">\r\n\r\nThe steering committee that selects the organizers of ML Day each year consists of Sham Kakade, Adam Kalai, and Joshua Tenenbaum.\r\n\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>\r\n<\/div>"}],"msr_startdate":"2014-05-13","msr_enddate":"2014-05-13","msr_event_time":"","msr_location":"Cambridge, MA, USA","msr_event_link":"","msr_event_recording_link":"","msr_startdate_formatted":"May 13, 2014","msr_register_text":"Watch now","msr_cta_link":"","msr_cta_text":"","msr_cta_bi_name":"","featured_image_thumbnail":null,"event_excerpt":"The third annual New England Machine Learning Day will be held May 13, 2014, at Microsoft Research New England, One Memorial Drive, Cambridge, MA 02142. The event will bring together local academics and researchers in machine learning and its applications.","msr_research_lab":[199563],"related-researchers":[],"msr_impact_theme":[],"related-academic-programs":[],"related-groups":[],"related-projects":[],"related-opportunities":[],"related-publications":[],"related-videos":[],"related-posts":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/199960","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-event"}],"version-history":[{"count":6,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/199960\/revisions"}],"predecessor-version":[{"id":1147386,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/199960\/revisions\/1147386"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=199960"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=199960"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=199960"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=199960"},{"taxonomy":"msr-video-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-video-type?post=199960"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=199960"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=199960"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=199960"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=199960"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}