{"id":394283,"date":"2017-07-03T16:08:24","date_gmt":"2017-07-03T23:08:24","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-event&#038;p=394283"},"modified":"2025-08-06T11:57:50","modified_gmt":"2025-08-06T18:57:50","slug":"international-conference-on-machine-learning-icml-2017","status":"publish","type":"msr-event","link":"https:\/\/www.microsoft.com\/en-us\/research\/event\/international-conference-on-machine-learning-icml-2017\/","title":{"rendered":"Microsoft Research @ ICML 2017"},"content":{"rendered":"\n\n<p><strong>Venue:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" title=\"map\" href=\"https:\/\/www.google.com\/maps\/place\/ICC+Sydney\/@-33.8733374,151.1992872,15z\/data=!4m5!3m4!1s0x0:0x19fc67061ce45a9!8m2!3d-33.8733374!4d151.1992872\" target=\"_blank\" rel=\"noopener\"><span lang=\"EN\">International Convention Centre<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p><strong>Co-located:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"http:\/\/auai.org\/uai2017\/index.php\" target=\"_blank\" rel=\"noopener\">Conference on Uncertainty in Artificial Intelligence (UAI)<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p><strong>Website:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/2017.icml.cc\/\" target=\"_blank\" rel=\"noopener\">International Conference on Machine Learning (ICML)\u00a02017<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>The International Conference on Machine Learning (ICML) is the leading international machine learning conference. Click <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/2017.icml.cc\/\" target=\"_blank\" rel=\"noopener\">here<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> to learn more.<\/p>\n<h2>Program committee members<\/h2>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lmackey\/\">Lester Mackey<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pkohli\/\">Pushmeet Kohli<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ryoto\/\">Ryota Tomioka<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"http:\/\/www.microsoft.com\/en-us\/research\/people\/sebubeck\/\">Sebastian Bubeck<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/senowozi\/\">Sebastian Nowozin<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<\/ul>\n<h2>Tutorials<\/h2>\n<ul>\n<li><span lang=\"EN\"><strong>Real World Interactive Learning<\/strong>, <\/span><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/alekhagarwal.net\/\"><span lang=\"EN\">Alekh Agarwal<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span lang=\"EN\"> (Microsoft Research) and <\/span><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/hunch.net\/~jl\"><span lang=\"EN\">John Langford<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span lang=\"EN\"> (Microsoft Research)<\/span><\/li>\n<li><span lang=\"EN\"><strong>Recent Advances in Stochastic Convex and Non-Convex Optimization<\/strong>, <\/span><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/people.csail.mit.edu\/zeyuan\/\"><span lang=\"EN\">Zeyuan Allen-Zhu<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span lang=\"EN\"> (Microsoft Research)<\/span><\/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><b>Doubly Accelerated Methods for Faster CCA and Generalized Eigendecomposition<br \/>\n<\/b>Zeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)<\/p>\n<p><b>Follow the Compressed Leader: Even Faster Online Learning of Eigenvectors<br \/>\n<\/b>Zeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)<\/p>\n<p><b>Faster Principal Component Regression via Optimal Polynomial Approximation to Matrix sgn(x)<\/b><br \/>\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)<\/p>\n<p><b>Sequence Modeling via Segmentations<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/chowang\/\">Chong Wang<\/a> (Microsoft Research) \u00b7 Yining Wang (CMU) \u00b7 Po-Sen Huang (Microsoft Research) \u00b7 Abdelrahman Mohammad (Microsoft) \u00b7 Dengyong Zhou (Microsoft Research) \u00b7 Li Deng (Citadel)<\/p>\n<p><b>Measuring Sample Quality with Kernels<\/b><br \/>\nJackson Gorham (STANFORD) \u00b7 Lester Mackey (Microsoft Research)<\/p>\n<p><b>Asynchronous Stochastic Gradient Descent with Delay Compensation<\/b><br \/>\nShuxin Zheng (University of Science and Technology of China) \u00b7 Qi Meng (Peking University) \u00b7 Taifeng Wang (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wche\/\">Wei Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a> (Microsoft)<\/p>\n<p><b>Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter<\/b><br \/>\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS)<\/p>\n<p><b>Near-Optimal Design of Experiments via Regret Minimization<\/b><br \/>\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University) \u00b7 Aarti Singh () \u00b7 Yining Wang (CMU)<\/p>\n<p><b>Contextual Decision Processes with low Bellman rank are PAC-Learnable<\/b><br \/>\nNan Jiang (Microsoft Research) \u00b7 Akshay Krishnamurthy (UMass) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal <\/a>(Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/schapire\/\">Robert Schapire<\/a> (Microsoft Research)<\/p>\n<p><b>Logarithmic Time One-Against-Some<\/b><br \/>\nHal Daum\u00e9 (University of Maryland) \u00b7 NIKOS KARAMPATZIAKIS (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford<\/a> (Microsoft Research) \u00b7 Paul Mineiro (Microsoft)<\/p>\n<p><b>Optimal and Adaptive Off-policy Evaluation in Contextual Bandits<\/b><br \/>\nYu-Xiang Wang (Carnegie Mellon University \/ Amazon AWS) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research) \u00b7 Miroslav Dudik (Microsoft Research)<\/p>\n<p><b>Safety-Aware Algorithms for Adversarial Contextual Bandit<\/b><br \/>\nWen Sun (Carnegie Mellon University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dedey\/\">Debadeepta Dey<\/a> (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/akapoor\/\">Ashish Kapoor<\/a> (Microsoft Research)<\/p>\n<p><b>How to Escape Saddle Points Efficiently<\/b><br \/>\nChi Jin (UC Berkeley) \u00b7 Rong Ge (Duke University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/praneeth\/\">Praneeth Netrapalli<\/a> (Microsoft Research) \u00b7 Sham M. Kakade (University of Washington) \u00b7 Michael Jordan ()<\/p>\n<p><b>Stochastic Variance Reduction Methods for Policy Evaluation<\/b><br \/>\nSimon Du (Carnegie Mellon University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jianshuc\/\">Jianshu Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lixiao\/\">Lin Xiao<\/a> (Microsoft Research) \u00b7 Dengyong Zhou (Microsoft Research)<\/p>\n<p><b>Provable Optimal Algorithms for Generalized Linear Contextual Bandits<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research) \u00b7 Yu Lu (Yale University) \u00b7 Dengyong Zhou (Microsoft Research)<\/p>\n<p><b>Learning Continuous Semantic Representations of Symbolic Expressions<\/b><br \/>\nMiltiadis Allamanis (Microsoft Research) \u00b7 pankajan Chanthirasegaran () \u00b7 Pushmeet Kohli (Microsoft Research) \u00b7 Charles Sutton (University of Edinburgh)<\/p>\n<p><b>RobustFill: Neural Program Learning under Noisy I\/O<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jdevlin\/\">Jacob Devlin<\/a> (Microsoft Research) \u00b7 Jonathan Uesato (MIT) \u00b7 Surya Bhupatiraju (MIT) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/risin\/\">Rishabh Singh<\/a> (Microsoft Research) \u00b7 Abdelrahman Mohammad (Microsoft) \u00b7 Pushmeet Kohli (Microsoft Research)<\/p>\n<p><b>Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks<\/b><br \/>\nLars Mescheder (MPI T\u00fcbingen) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/senowozi\/\">Sebastian Nowozin<\/a> (Microsoft Research) \u00b7 Andreas Geiger (MPI T\u00fcbingen)<\/p>\n<p><b>ProtoNN: Compressed and Accurate kNN for Resource-scarce Devices<\/b><br \/>\nChirag Gupta (Microsoft Research, India) \u00b7 ARUN SUGGALA (Carnegie Mellon University) \u00b7 Ankit Goyal (University of Michigan) \u00b7 Saurabh Goyal (IBM India Pvt Ltd) \u00b7 Ashish Kumar (Microsoft Research) \u00b7 Bhargavi Paranjape (Microsoft Research) \u00b7 Harsha Vardhan Simhadri (Microsoft Research) \u00b7 Raghavendra Udupa (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/manik\/\">Manik Varma<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research)<\/p>\n<p><b>Optimal algorithms for smooth and strongly convex distributed optimization in networks<\/b><br \/>\nKevin Scaman (MSR-INRIA Joint Center) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/yile\/\">Yin Tat Lee<\/a> (Microsoft Research) \u00b7 Francis Bach (INRIA) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sebubeck\/\">Sebastien Bubeck<\/a> (Microsoft Research) \u00b7 Laurent Massouli\u00e9 (MSR-INRIA Joint Center)<\/p>\n<p><b>Resource-efficient Machine Learning in 2 KB RAM for the Internet of Things<\/b><br \/>\nAshish Kumar (Microsoft Research) \u00b7 Saurabh Goyal (IBM India Pvt Ltd) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/manik\/\">Manik Varma<\/a> (Microsoft Research)<\/p>\n<p><b>Batched High-dimensional Bayesian Optimization via Structural Kernel Learning<\/b><br \/>\nZi Wang (MIT) \u00b7 Chengtao Li () \u00b7 Stefanie Jegelka (MIT) \u00b7 Pushmeet Kohli (Microsoft Research)<\/p>\n<p><b>Recovery Guarantees for One-hidden-layer Neural Networks<\/b><br \/>\nKai Zhong (University of Texas at Austin) \u00b7 Zhao Song (UT-Austin) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 Peter Bartlett (UC Berkeley) \u00b7 Inderjit Dhillon (UT Austin & Amazon)<\/p>\n<p><b>Dual Supervised Learning<\/b><br \/>\nYingce Xia (University of Science and Technology of China) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a> (Microsoft Research Asia) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wche\/\">Wei Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jibian\/\">Jiang Bian<\/a> (Microsoft Research) \u00b7 Nenghai Yu (USTC) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a> (Microsoft)<\/p>\n<p><b>Improving Gibbs Sampler Scan Quality with DoGS<\/b><br \/>\nIoannis Mitliagkas (Stanford University) \u00b7 Lester Mackey (Microsoft Research)<\/p>\n<p><b>Nearly Optimal Robust Matrix Completion<\/b><br \/>\nYeshwanth Cherapanamjeri (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 Kartik Gupta (Microsoft Research)<\/p>\n<p><b>Stabilising Experience Replay for Deep Multi-Agent Reinforcement Learning<\/b><br \/>\nJakob Foerster (University of Oxford) \u00b7 Nantas Nardelli (University of Oxford) \u00b7 Gregory Farquhar (University of Oxford) \u00b7 Phil Torr (Oxford) \u00b7 Pushmeet Kohli (Microsoft Research) \u00b7 Shimon Whiteson (University of Oxford)<\/p>\n<p><b>Differentiable Programs with Neural Libraries<\/b><br \/>\nAlex Gaunt (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mabrocks\/\">Marc Brockschmidt<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nkushman\/\">Nate Kushman<\/a> (Microsoft Research) \u00b7 Daniel Tarlow (Google Brain)<\/p>\n<p><b>Active Heteroscedastic Regression<\/b><br \/>\nKamalika Chaudhuri (University of California at San Diego) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/t-nanata\/\">Nagarajan Natarajan<\/a> (Microsoft Research)<\/p>\n<p><b>Consistency Analysis for Binary Classification Revisited<\/b><br \/>\nWojciech Kotlowski (Poznan University of Technology) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/t-nanata\/\">Nagarajan Natarajan<\/a> (Microsoft Research) \u00b7 Krzysztof Dembczynski (Poznan University of Technology) \u00b7 Oluwasanmi Koyejo (University of Illinois at Urbana-Champaign)<\/p>\n<p><b>Active Learning for Cost-Sensitive Classification<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research) \u00b7 Akshay Krishnamurthy (UMass) \u00b7 Tzu-Kuo Huang (Uber) \u00b7 Hal Daum\u00e9 III (University of Maryland) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford <\/a>(Microsoft Research)<\/p>\n<p><b>Adaptive Neural Networks for Fast Test-Time Prediction<\/b><br \/>\nTolga Bolukbasi (Boston University) \u00b7 Joseph Wang (Amazon) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/oferd\/\">Ofer Dekel<\/a> (Microsoft) \u00b7 Venkatesh Saligrama (Boston University)<\/p>\n<p><b>Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning<\/b><br \/>\nJunhyuk Oh (University of Michigan) \u00b7 Satinder Singh (University of Michigan) \u00b7 Honglak Lee (Google \/ U. Michigan) \u00b7 Pushmeet Kohli (Microsoft Research)<\/p>\n<p><b>Robust Structured Estimation with Single-Index Models<\/b><br \/>\nSheng Chen (University of Minnesota) \u00b7 Arindam Banerjee (University of Minnesota) \u00b7 Sreangsu Acharyya (Microsoft Research India)<\/p>\n<p><b>Gradient Coding: Avoiding Stragglers in Distributed Learning<\/b><br \/>\nRashish Tandon (University of Texas at Austin) \u00b7 Qi Lei (University of Texas at Austin) \u00b7 Alexandros Dimakis (UT Austin) \u00b7 NIKOS KARAMPATZIAKIS (Microsoft)<\/p>\n<p><b>Exploiting Strong Convexity from Data with Primal-Dual First-Order Algorithms<\/b><br \/>\nJialei Wang (University of Chicago) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lixiao\/\">Lin Xiao<\/a> (Microsoft Research)<\/p>\n<p><b>Gradient Boosted Decision Trees for High Dimensional Sparse Output<\/b><br \/>\nSi Si (Google Research) \u00b7 Huan Zhang (UC Davis) \u00b7 Sathiya Keerthi (Microsoft) \u00b7 Dhruv Mahajan (Facebook) \u00b7 Inderjit Dhillon (UT Austin & Amazon) \u00b7 Cho-Jui Hsieh (University of California, Davis)<\/p>\n<p><b>Learning Algorithms for Active Learning<\/b><br \/>\nPhilip Bachman (Maluuba) \u00b7 Alessandro Sordoni (Microsoft Maluuba) \u00b7 Adam Trischler (Maluuba)<\/p>\n<p><b>Deep IV: A Flexible Approach for Counterfactual Prediction<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/glewis\/\">Greg Lewis<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taddy\/\">Matt Taddy<\/a> (MICROSOFT) \u00b7 Jason Hartford (University of British Columbia) \u00b7 Kevin Leyton-Brown ()<span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Venue: International Convention Centre (opens in new tab) Co-located: Conference on Uncertainty in Artificial Intelligence (UAI) (opens in new tab) Website: International Conference on Machine Learning (ICML)\u00a02017 (opens in new tab)Opens in a new tab The International Conference on Machine Learning (ICML) is the leading international machine learning conference. Click here (opens in new tab) [&hellip;]<\/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":"2017-08-06","msr_enddate":"2017-08-11","msr_location":"Sydney, Australia","msr_expirationdate":"","msr_event_recording_link":"","msr_event_link":"","msr_event_link_redirect":false,"msr_event_time":"","msr_hide_region":false,"msr_private_event":false,"msr_hide_image_in_river":0,"footnotes":""},"research-area":[13556],"msr-region":[197900],"msr-event-type":[197941],"msr-video-type":[],"msr-locale":[268875],"msr-program-audience":[],"msr-post-option":[],"msr-impact-theme":[],"class_list":["post-394283","msr-event","type-msr-event","status-publish","hentry","msr-research-area-artificial-intelligence","msr-region-north-america","msr-event-type-conferences","msr-locale-en_us"],"msr_about":"<!-- wp:msr\/event-details {\"title\":\"Microsoft Research @ ICML 2017\",\"backgroundColor\":\"grey\"} \/-->\n\n<!-- wp:msr\/content-tabs --><!-- wp:msr\/content-tab {\"title\":\"About\"} --><!-- wp:freeform --><p><strong>Venue:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" title=\"map\" href=\"https:\/\/www.google.com\/maps\/place\/ICC+Sydney\/@-33.8733374,151.1992872,15z\/data=!4m5!3m4!1s0x0:0x19fc67061ce45a9!8m2!3d-33.8733374!4d151.1992872\" target=\"_blank\" rel=\"noopener\"><span lang=\"EN\">International Convention Centre<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p><strong>Co-located:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"http:\/\/auai.org\/uai2017\/index.php\" target=\"_blank\" rel=\"noopener\">Conference on Uncertainty in Artificial Intelligence (UAI)<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><\/p>\n<p><strong>Website:<\/strong> <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/2017.icml.cc\/\" target=\"_blank\" rel=\"noopener\">International Conference on Machine Learning (ICML)\u00a02017<span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span id=\"label-external-link\" class=\"sr-only\" aria-hidden=\"true\">Opens in a new tab<\/span><\/p>\n<p>The International Conference on Machine Learning (ICML) is the leading international machine learning conference. Click <a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" href=\"https:\/\/2017.icml.cc\/\" target=\"_blank\" rel=\"noopener\">here<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> to learn more.<\/p>\n<h2>Program committee members<\/h2>\n<ul>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lmackey\/\">Lester Mackey<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pkohli\/\">Pushmeet Kohli<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ryoto\/\">Ryota Tomioka<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"http:\/\/www.microsoft.com\/en-us\/research\/people\/sebubeck\/\">Sebastian Bubeck<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/senowozi\/\">Sebastian Nowozin<span class=\"sr-only\"> (opens in new tab)<\/span><\/a> (Microsoft Research)<\/li>\n<\/ul>\n<h2>Tutorials<\/h2>\n<ul>\n<li><span lang=\"EN\"><strong>Real World Interactive Learning<\/strong>, <\/span><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/alekhagarwal.net\/\"><span lang=\"EN\">Alekh Agarwal<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span lang=\"EN\"> (Microsoft Research) and <\/span><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/hunch.net\/~jl\"><span lang=\"EN\">John Langford<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span lang=\"EN\"> (Microsoft Research)<\/span><\/li>\n<li><span lang=\"EN\"><strong>Recent Advances in Stochastic Convex and Non-Convex Optimization<\/strong>, <\/span><a class=\"msr-external-link glyph-append glyph-append-open-in-new-tab glyph-append-xsmall\" target=\"_blank\" href=\"http:\/\/people.csail.mit.edu\/zeyuan\/\"><span lang=\"EN\">Zeyuan Allen-Zhu<\/span><span class=\"sr-only\"> (opens in new tab)<\/span><\/a><span lang=\"EN\"> (Microsoft Research)<\/span><\/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\":\"Accepted Papers\"} --><!-- wp:freeform --><p><b>Doubly Accelerated Methods for Faster CCA and Generalized Eigendecomposition<br \/>\n<\/b>Zeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)<\/p>\n<p><b>Follow the Compressed Leader: Even Faster Online Learning of Eigenvectors<br \/>\n<\/b>Zeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)<\/p>\n<p><b>Faster Principal Component Regression via Optimal Polynomial Approximation to Matrix sgn(x)<\/b><br \/>\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)<\/p>\n<p><b>Sequence Modeling via Segmentations<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/chowang\/\">Chong Wang<\/a> (Microsoft Research) \u00b7 Yining Wang (CMU) \u00b7 Po-Sen Huang (Microsoft Research) \u00b7 Abdelrahman Mohammad (Microsoft) \u00b7 Dengyong Zhou (Microsoft Research) \u00b7 Li Deng (Citadel)<\/p>\n<p><b>Measuring Sample Quality with Kernels<\/b><br \/>\nJackson Gorham (STANFORD) \u00b7 Lester Mackey (Microsoft Research)<\/p>\n<p><b>Asynchronous Stochastic Gradient Descent with Delay Compensation<\/b><br \/>\nShuxin Zheng (University of Science and Technology of China) \u00b7 Qi Meng (Peking University) \u00b7 Taifeng Wang (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wche\/\">Wei Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a> (Microsoft)<\/p>\n<p><b>Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter<\/b><br \/>\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS)<\/p>\n<p><b>Near-Optimal Design of Experiments via Regret Minimization<\/b><br \/>\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University) \u00b7 Aarti Singh () \u00b7 Yining Wang (CMU)<\/p>\n<p><b>Contextual Decision Processes with low Bellman rank are PAC-Learnable<\/b><br \/>\nNan Jiang (Microsoft Research) \u00b7 Akshay Krishnamurthy (UMass) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal <\/a>(Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/schapire\/\">Robert Schapire<\/a> (Microsoft Research)<\/p>\n<p><b>Logarithmic Time One-Against-Some<\/b><br \/>\nHal Daum\u00e9 (University of Maryland) \u00b7 NIKOS KARAMPATZIAKIS (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford<\/a> (Microsoft Research) \u00b7 Paul Mineiro (Microsoft)<\/p>\n<p><b>Optimal and Adaptive Off-policy Evaluation in Contextual Bandits<\/b><br \/>\nYu-Xiang Wang (Carnegie Mellon University \/ Amazon AWS) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research) \u00b7 Miroslav Dudik (Microsoft Research)<\/p>\n<p><b>Safety-Aware Algorithms for Adversarial Contextual Bandit<\/b><br \/>\nWen Sun (Carnegie Mellon University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dedey\/\">Debadeepta Dey<\/a> (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/akapoor\/\">Ashish Kapoor<\/a> (Microsoft Research)<\/p>\n<p><b>How to Escape Saddle Points Efficiently<\/b><br \/>\nChi Jin (UC Berkeley) \u00b7 Rong Ge (Duke University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/praneeth\/\">Praneeth Netrapalli<\/a> (Microsoft Research) \u00b7 Sham M. Kakade (University of Washington) \u00b7 Michael Jordan ()<\/p>\n<p><b>Stochastic Variance Reduction Methods for Policy Evaluation<\/b><br \/>\nSimon Du (Carnegie Mellon University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jianshuc\/\">Jianshu Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lixiao\/\">Lin Xiao<\/a> (Microsoft Research) \u00b7 Dengyong Zhou (Microsoft Research)<\/p>\n<p><b>Provable Optimal Algorithms for Generalized Linear Contextual Bandits<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research) \u00b7 Yu Lu (Yale University) \u00b7 Dengyong Zhou (Microsoft Research)<\/p>\n<p><b>Learning Continuous Semantic Representations of Symbolic Expressions<\/b><br \/>\nMiltiadis Allamanis (Microsoft Research) \u00b7 pankajan Chanthirasegaran () \u00b7 Pushmeet Kohli (Microsoft Research) \u00b7 Charles Sutton (University of Edinburgh)<\/p>\n<p><b>RobustFill: Neural Program Learning under Noisy I\/O<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jdevlin\/\">Jacob Devlin<\/a> (Microsoft Research) \u00b7 Jonathan Uesato (MIT) \u00b7 Surya Bhupatiraju (MIT) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/risin\/\">Rishabh Singh<\/a> (Microsoft Research) \u00b7 Abdelrahman Mohammad (Microsoft) \u00b7 Pushmeet Kohli (Microsoft Research)<\/p>\n<p><b>Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks<\/b><br \/>\nLars Mescheder (MPI T\u00fcbingen) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/senowozi\/\">Sebastian Nowozin<\/a> (Microsoft Research) \u00b7 Andreas Geiger (MPI T\u00fcbingen)<\/p>\n<p><b>ProtoNN: Compressed and Accurate kNN for Resource-scarce Devices<\/b><br \/>\nChirag Gupta (Microsoft Research, India) \u00b7 ARUN SUGGALA (Carnegie Mellon University) \u00b7 Ankit Goyal (University of Michigan) \u00b7 Saurabh Goyal (IBM India Pvt Ltd) \u00b7 Ashish Kumar (Microsoft Research) \u00b7 Bhargavi Paranjape (Microsoft Research) \u00b7 Harsha Vardhan Simhadri (Microsoft Research) \u00b7 Raghavendra Udupa (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/manik\/\">Manik Varma<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research)<\/p>\n<p><b>Optimal algorithms for smooth and strongly convex distributed optimization in networks<\/b><br \/>\nKevin Scaman (MSR-INRIA Joint Center) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/yile\/\">Yin Tat Lee<\/a> (Microsoft Research) \u00b7 Francis Bach (INRIA) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sebubeck\/\">Sebastien Bubeck<\/a> (Microsoft Research) \u00b7 Laurent Massouli\u00e9 (MSR-INRIA Joint Center)<\/p>\n<p><b>Resource-efficient Machine Learning in 2 KB RAM for the Internet of Things<\/b><br \/>\nAshish Kumar (Microsoft Research) \u00b7 Saurabh Goyal (IBM India Pvt Ltd) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/manik\/\">Manik Varma<\/a> (Microsoft Research)<\/p>\n<p><b>Batched High-dimensional Bayesian Optimization via Structural Kernel Learning<\/b><br \/>\nZi Wang (MIT) \u00b7 Chengtao Li () \u00b7 Stefanie Jegelka (MIT) \u00b7 Pushmeet Kohli (Microsoft Research)<\/p>\n<p><b>Recovery Guarantees for One-hidden-layer Neural Networks<\/b><br \/>\nKai Zhong (University of Texas at Austin) \u00b7 Zhao Song (UT-Austin) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 Peter Bartlett (UC Berkeley) \u00b7 Inderjit Dhillon (UT Austin &amp; Amazon)<\/p>\n<p><b>Dual Supervised Learning<\/b><br \/>\nYingce Xia (University of Science and Technology of China) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a> (Microsoft Research Asia) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wche\/\">Wei Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jibian\/\">Jiang Bian<\/a> (Microsoft Research) \u00b7 Nenghai Yu (USTC) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a> (Microsoft)<\/p>\n<p><b>Improving Gibbs Sampler Scan Quality with DoGS<\/b><br \/>\nIoannis Mitliagkas (Stanford University) \u00b7 Lester Mackey (Microsoft Research)<\/p>\n<p><b>Nearly Optimal Robust Matrix Completion<\/b><br \/>\nYeshwanth Cherapanamjeri (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 Kartik Gupta (Microsoft Research)<\/p>\n<p><b>Stabilising Experience Replay for Deep Multi-Agent Reinforcement Learning<\/b><br \/>\nJakob Foerster (University of Oxford) \u00b7 Nantas Nardelli (University of Oxford) \u00b7 Gregory Farquhar (University of Oxford) \u00b7 Phil Torr (Oxford) \u00b7 Pushmeet Kohli (Microsoft Research) \u00b7 Shimon Whiteson (University of Oxford)<\/p>\n<p><b>Differentiable Programs with Neural Libraries<\/b><br \/>\nAlex Gaunt (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mabrocks\/\">Marc Brockschmidt<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nkushman\/\">Nate Kushman<\/a> (Microsoft Research) \u00b7 Daniel Tarlow (Google Brain)<\/p>\n<p><b>Active Heteroscedastic Regression<\/b><br \/>\nKamalika Chaudhuri (University of California at San Diego) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/t-nanata\/\">Nagarajan Natarajan<\/a> (Microsoft Research)<\/p>\n<p><b>Consistency Analysis for Binary Classification Revisited<\/b><br \/>\nWojciech Kotlowski (Poznan University of Technology) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/t-nanata\/\">Nagarajan Natarajan<\/a> (Microsoft Research) \u00b7 Krzysztof Dembczynski (Poznan University of Technology) \u00b7 Oluwasanmi Koyejo (University of Illinois at Urbana-Champaign)<\/p>\n<p><b>Active Learning for Cost-Sensitive Classification<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research) \u00b7 Akshay Krishnamurthy (UMass) \u00b7 Tzu-Kuo Huang (Uber) \u00b7 Hal Daum\u00e9 III (University of Maryland) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford <\/a>(Microsoft Research)<\/p>\n<p><b>Adaptive Neural Networks for Fast Test-Time Prediction<\/b><br \/>\nTolga Bolukbasi (Boston University) \u00b7 Joseph Wang (Amazon) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/oferd\/\">Ofer Dekel<\/a> (Microsoft) \u00b7 Venkatesh Saligrama (Boston University)<\/p>\n<p><b>Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning<\/b><br \/>\nJunhyuk Oh (University of Michigan) \u00b7 Satinder Singh (University of Michigan) \u00b7 Honglak Lee (Google \/ U. Michigan) \u00b7 Pushmeet Kohli (Microsoft Research)<\/p>\n<p><b>Robust Structured Estimation with Single-Index Models<\/b><br \/>\nSheng Chen (University of Minnesota) \u00b7 Arindam Banerjee (University of Minnesota) \u00b7 Sreangsu Acharyya (Microsoft Research India)<\/p>\n<p><b>Gradient Coding: Avoiding Stragglers in Distributed Learning<\/b><br \/>\nRashish Tandon (University of Texas at Austin) \u00b7 Qi Lei (University of Texas at Austin) \u00b7 Alexandros Dimakis (UT Austin) \u00b7 NIKOS KARAMPATZIAKIS (Microsoft)<\/p>\n<p><b>Exploiting Strong Convexity from Data with Primal-Dual First-Order Algorithms<\/b><br \/>\nJialei Wang (University of Chicago) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lixiao\/\">Lin Xiao<\/a> (Microsoft Research)<\/p>\n<p><b>Gradient Boosted Decision Trees for High Dimensional Sparse Output<\/b><br \/>\nSi Si (Google Research) \u00b7 Huan Zhang (UC Davis) \u00b7 Sathiya Keerthi (Microsoft) \u00b7 Dhruv Mahajan (Facebook) \u00b7 Inderjit Dhillon (UT Austin &amp; Amazon) \u00b7 Cho-Jui Hsieh (University of California, Davis)<\/p>\n<p><b>Learning Algorithms for Active Learning<\/b><br \/>\nPhilip Bachman (Maluuba) \u00b7 Alessandro Sordoni (Microsoft Maluuba) \u00b7 Adam Trischler (Maluuba)<\/p>\n<p><b>Deep IV: A Flexible Approach for Counterfactual Prediction<\/b><br \/>\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/glewis\/\">Greg Lewis<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taddy\/\">Matt Taddy<\/a> (MICROSOFT) \u00b7 Jason Hartford (University of British Columbia) \u00b7 Kevin Leyton-Brown ()<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 International Conference on Machine Learning (ICML) is the leading international machine learning conference. Click <a href=\"https:\/\/2017.icml.cc\/\" target=\"_blank\" rel=\"noopener\">here<\/a> to learn more.\r\n<h2>Program committee members<\/h2>\r\n<ul>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research)<\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lmackey\/\">Lester Mackey<\/a> (Microsoft Research)<\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research)<\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/pkohli\/\">Pushmeet Kohli<\/a> (Microsoft Research)<\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/ryoto\/\">Ryota Tomioka<\/a> (Microsoft Research)<\/li>\r\n \t<li><a href=\"http:\/\/www.microsoft.com\/en-us\/research\/people\/sebubeck\/\">Sebastian Bubeck<\/a> (Microsoft Research)<\/li>\r\n \t<li><a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/senowozi\/\">Sebastian Nowozin<\/a> (Microsoft Research)<\/li>\r\n<\/ul>\r\n<h2>Tutorials<\/h2>\r\n<ul>\r\n \t<li><span lang=\"EN\"><strong>Real World Interactive Learning<\/strong>, <\/span><a href=\"http:\/\/alekhagarwal.net\/\"><span lang=\"EN\">Alekh Agarwal<\/span><\/a><span lang=\"EN\"> (Microsoft Research) and <\/span><a href=\"http:\/\/hunch.net\/~jl\"><span lang=\"EN\">John Langford<\/span><\/a><span lang=\"EN\"> (Microsoft Research)<\/span><\/li>\r\n \t<li><span lang=\"EN\"><strong>Recent Advances in Stochastic Convex and Non-Convex Optimization<\/strong>, <\/span><a href=\"http:\/\/people.csail.mit.edu\/zeyuan\/\"><span lang=\"EN\">Zeyuan Allen-Zhu<\/span><\/a><span lang=\"EN\"> (Microsoft Research)<\/span><\/li>\r\n<\/ul>"},{"id":1,"name":"Accepted Papers","content":"<b>Doubly Accelerated Methods for Faster CCA and Generalized Eigendecomposition\r\n<\/b>Zeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)\r\n\r\n<b>Follow the Compressed Leader: Even Faster Online Learning of Eigenvectors\r\n<\/b>Zeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)\r\n\r\n<b>Faster Principal Component Regression via Optimal Polynomial Approximation to Matrix sgn(x)<\/b>\r\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University)\r\n\r\n<b>Sequence Modeling via Segmentations<\/b>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/chowang\/\">Chong Wang<\/a> (Microsoft Research) \u00b7 Yining Wang (CMU) \u00b7 Po-Sen Huang (Microsoft Research) \u00b7 Abdelrahman Mohammad (Microsoft) \u00b7 Dengyong Zhou (Microsoft Research) \u00b7 Li Deng (Citadel)\r\n\r\n<b>Measuring Sample Quality with Kernels<\/b>\r\nJackson Gorham (STANFORD) \u00b7 Lester Mackey (Microsoft Research)\r\n\r\n<b>Asynchronous Stochastic Gradient Descent with Delay Compensation<\/b>\r\nShuxin Zheng (University of Science and Technology of China) \u00b7 Qi Meng (Peking University) \u00b7 Taifeng Wang (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wche\/\">Wei Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a> (Microsoft)\r\n\r\n<b>Natasha: Faster Non-Convex Stochastic Optimization Via Strongly Non-Convex Parameter<\/b>\r\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS)\r\n\r\n<b>Near-Optimal Design of Experiments via Regret Minimization<\/b>\r\nZeyuan Allen-Zhu (Microsoft Research \/ Princeton \/ IAS) \u00b7 Yuanzhi Li (Princeton University) \u00b7 Aarti Singh () \u00b7 Yining Wang (CMU)\r\n\r\n<b>Contextual Decision Processes with low Bellman rank are PAC-Learnable<\/b>\r\nNan Jiang (Microsoft Research) \u00b7 Akshay Krishnamurthy (UMass) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal <\/a>(Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/schapire\/\">Robert Schapire<\/a> (Microsoft Research)\r\n\r\n<b>Logarithmic Time One-Against-Some<\/b>\r\nHal Daum\u00e9 (University of Maryland) \u00b7 NIKOS KARAMPATZIAKIS (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford<\/a> (Microsoft Research) \u00b7 Paul Mineiro (Microsoft)\r\n\r\n<b>Optimal and Adaptive Off-policy Evaluation in Contextual Bandits<\/b>\r\nYu-Xiang Wang (Carnegie Mellon University \/ Amazon AWS) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research) \u00b7 Miroslav Dudik (Microsoft Research)\r\n\r\n<b>Safety-Aware Algorithms for Adversarial Contextual Bandit<\/b>\r\nWen Sun (Carnegie Mellon University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/dedey\/\">Debadeepta Dey<\/a> (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/akapoor\/\">Ashish Kapoor<\/a> (Microsoft Research)\r\n\r\n<b>How to Escape Saddle Points Efficiently<\/b>\r\nChi Jin (UC Berkeley) \u00b7 Rong Ge (Duke University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/praneeth\/\">Praneeth Netrapalli<\/a> (Microsoft Research) \u00b7 Sham M. Kakade (University of Washington) \u00b7 Michael Jordan ()\r\n\r\n<b>Stochastic Variance Reduction Methods for Policy Evaluation<\/b>\r\nSimon Du (Carnegie Mellon University) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jianshuc\/\">Jianshu Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lixiao\/\">Lin Xiao<\/a> (Microsoft Research) \u00b7 Dengyong Zhou (Microsoft Research)\r\n\r\n<b>Provable Optimal Algorithms for Generalized Linear Contextual Bandits<\/b>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lihongli\/\">Lihong Li<\/a> (Microsoft Research) \u00b7 Yu Lu (Yale University) \u00b7 Dengyong Zhou (Microsoft Research)\r\n\r\n<b>Learning Continuous Semantic Representations of Symbolic Expressions<\/b>\r\nMiltiadis Allamanis (Microsoft Research) \u00b7 pankajan Chanthirasegaran () \u00b7 Pushmeet Kohli (Microsoft Research) \u00b7 Charles Sutton (University of Edinburgh)\r\n\r\n<b>RobustFill: Neural Program Learning under Noisy I\/O<\/b>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jdevlin\/\">Jacob Devlin<\/a> (Microsoft Research) \u00b7 Jonathan Uesato (MIT) \u00b7 Surya Bhupatiraju (MIT) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/risin\/\">Rishabh Singh<\/a> (Microsoft Research) \u00b7 Abdelrahman Mohammad (Microsoft) \u00b7 Pushmeet Kohli (Microsoft Research)\r\n\r\n<b>Adversarial Variational Bayes: Unifying Variational Autoencoders and Generative Adversarial Networks<\/b>\r\nLars Mescheder (MPI T\u00fcbingen) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/senowozi\/\">Sebastian Nowozin<\/a> (Microsoft Research) \u00b7 Andreas Geiger (MPI T\u00fcbingen)\r\n\r\n<b>ProtoNN: Compressed and Accurate kNN for Resource-scarce Devices<\/b>\r\nChirag Gupta (Microsoft Research, India) \u00b7 ARUN SUGGALA (Carnegie Mellon University) \u00b7 Ankit Goyal (University of Michigan) \u00b7 Saurabh Goyal (IBM India Pvt Ltd) \u00b7 Ashish Kumar (Microsoft Research) \u00b7 Bhargavi Paranjape (Microsoft Research) \u00b7 Harsha Vardhan Simhadri (Microsoft Research) \u00b7 Raghavendra Udupa (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/manik\/\">Manik Varma<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research)\r\n\r\n<b>Optimal algorithms for smooth and strongly convex distributed optimization in networks<\/b>\r\nKevin Scaman (MSR-INRIA Joint Center) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/yile\/\">Yin Tat Lee<\/a> (Microsoft Research) \u00b7 Francis Bach (INRIA) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/sebubeck\/\">Sebastien Bubeck<\/a> (Microsoft Research) \u00b7 Laurent Massouli\u00e9 (MSR-INRIA Joint Center)\r\n\r\n<b>Resource-efficient Machine Learning in 2 KB RAM for the Internet of Things<\/b>\r\nAshish Kumar (Microsoft Research) \u00b7 Saurabh Goyal (IBM India Pvt Ltd) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/manik\/\">Manik Varma<\/a> (Microsoft Research)\r\n\r\n<b>Batched High-dimensional Bayesian Optimization via Structural Kernel Learning<\/b>\r\nZi Wang (MIT) \u00b7 Chengtao Li () \u00b7 Stefanie Jegelka (MIT) \u00b7 Pushmeet Kohli (Microsoft Research)\r\n\r\n<b>Recovery Guarantees for One-hidden-layer Neural Networks<\/b>\r\nKai Zhong (University of Texas at Austin) \u00b7 Zhao Song (UT-Austin) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 Peter Bartlett (UC Berkeley) \u00b7 Inderjit Dhillon (UT Austin &amp; Amazon)\r\n\r\n<b>Dual Supervised Learning<\/b>\r\nYingce Xia (University of Science and Technology of China) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taoqin\/\">Tao Qin<\/a> (Microsoft Research Asia) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/wche\/\">Wei Chen<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jibian\/\">Jiang Bian<\/a> (Microsoft Research) \u00b7 Nenghai Yu (USTC) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/tyliu\/\">Tie-Yan Liu<\/a> (Microsoft)\r\n\r\n<b>Improving Gibbs Sampler Scan Quality with DoGS<\/b>\r\nIoannis Mitliagkas (Stanford University) \u00b7 Lester Mackey (Microsoft Research)\r\n\r\n<b>Nearly Optimal Robust Matrix Completion<\/b>\r\nYeshwanth Cherapanamjeri (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 Kartik Gupta (Microsoft Research)\r\n\r\n<b>Stabilising Experience Replay for Deep Multi-Agent Reinforcement Learning<\/b>\r\nJakob Foerster (University of Oxford) \u00b7 Nantas Nardelli (University of Oxford) \u00b7 Gregory Farquhar (University of Oxford) \u00b7 Phil Torr (Oxford) \u00b7 Pushmeet Kohli (Microsoft Research) \u00b7 Shimon Whiteson (University of Oxford)\r\n\r\n<b>Differentiable Programs with Neural Libraries<\/b>\r\nAlex Gaunt (Microsoft) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/mabrocks\/\">Marc Brockschmidt<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/nkushman\/\">Nate Kushman<\/a> (Microsoft Research) \u00b7 Daniel Tarlow (Google Brain)\r\n\r\n<b>Active Heteroscedastic Regression<\/b>\r\nKamalika Chaudhuri (University of California at San Diego) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/prajain\/\">Prateek Jain<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/t-nanata\/\">Nagarajan Natarajan<\/a> (Microsoft Research)\r\n\r\n<b>Consistency Analysis for Binary Classification Revisited<\/b>\r\nWojciech Kotlowski (Poznan University of Technology) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/t-nanata\/\">Nagarajan Natarajan<\/a> (Microsoft Research) \u00b7 Krzysztof Dembczynski (Poznan University of Technology) \u00b7 Oluwasanmi Koyejo (University of Illinois at Urbana-Champaign)\r\n\r\n<b>Active Learning for Cost-Sensitive Classification<\/b>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/alekha\/\">Alekh Agarwal<\/a> (Microsoft Research) \u00b7 Akshay Krishnamurthy (UMass) \u00b7 Tzu-Kuo Huang (Uber) \u00b7 Hal Daum\u00e9 III (University of Maryland) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/jcl\/\">John Langford <\/a>(Microsoft Research)\r\n\r\n<b>Adaptive Neural Networks for Fast Test-Time Prediction<\/b>\r\nTolga Bolukbasi (Boston University) \u00b7 Joseph Wang (Amazon) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/oferd\/\">Ofer Dekel<\/a> (Microsoft) \u00b7 Venkatesh Saligrama (Boston University)\r\n\r\n<b>Zero-Shot Task Generalization with Multi-Task Deep Reinforcement Learning<\/b>\r\nJunhyuk Oh (University of Michigan) \u00b7 Satinder Singh (University of Michigan) \u00b7 Honglak Lee (Google \/ U. Michigan) \u00b7 Pushmeet Kohli (Microsoft Research)\r\n\r\n<b>Robust Structured Estimation with Single-Index Models<\/b>\r\nSheng Chen (University of Minnesota) \u00b7 Arindam Banerjee (University of Minnesota) \u00b7 Sreangsu Acharyya (Microsoft Research India)\r\n\r\n<b>Gradient Coding: Avoiding Stragglers in Distributed Learning<\/b>\r\nRashish Tandon (University of Texas at Austin) \u00b7 Qi Lei (University of Texas at Austin) \u00b7 Alexandros Dimakis (UT Austin) \u00b7 NIKOS KARAMPATZIAKIS (Microsoft)\r\n\r\n<b>Exploiting Strong Convexity from Data with Primal-Dual First-Order Algorithms<\/b>\r\nJialei Wang (University of Chicago) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/lixiao\/\">Lin Xiao<\/a> (Microsoft Research)\r\n\r\n<b>Gradient Boosted Decision Trees for High Dimensional Sparse Output<\/b>\r\nSi Si (Google Research) \u00b7 Huan Zhang (UC Davis) \u00b7 Sathiya Keerthi (Microsoft) \u00b7 Dhruv Mahajan (Facebook) \u00b7 Inderjit Dhillon (UT Austin &amp; Amazon) \u00b7 Cho-Jui Hsieh (University of California, Davis)\r\n\r\n<b>Learning Algorithms for Active Learning<\/b>\r\nPhilip Bachman (Maluuba) \u00b7 Alessandro Sordoni (Microsoft Maluuba) \u00b7 Adam Trischler (Maluuba)\r\n\r\n<b>Deep IV: A Flexible Approach for Counterfactual Prediction<\/b>\r\n<a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/glewis\/\">Greg Lewis<\/a> (Microsoft Research) \u00b7 <a href=\"https:\/\/www.microsoft.com\/en-us\/research\/people\/taddy\/\">Matt Taddy<\/a> (MICROSOFT) \u00b7 Jason Hartford (University of British Columbia) \u00b7 Kevin Leyton-Brown ()"}],"msr_startdate":"2017-08-06","msr_enddate":"2017-08-11","msr_event_time":"","msr_location":"Sydney, Australia","msr_event_link":"","msr_event_recording_link":"","msr_startdate_formatted":"August 6, 2017","msr_register_text":"Watch now","msr_cta_link":"","msr_cta_text":"","msr_cta_bi_name":"","featured_image_thumbnail":null,"event_excerpt":"The International Conference on Machine Learning (ICML) is the leading international machine learning conference. Click here to learn more. Program committee members Alekh Agarwal (Microsoft Research) Lester Mackey (Microsoft Research) Lihong Li (Microsoft Research) Pushmeet Kohli (Microsoft Research) Ryota Tomioka (Microsoft Research) Sebastian Bubeck (Microsoft Research) Sebastian Nowozin (Microsoft Research) Tutorials Real World Interactive Learning, Alekh Agarwal (Microsoft Research) and John Langford (Microsoft Research) Recent Advances in Stochastic Convex and Non-Convex Optimization, Zeyuan Allen-Zhu (Microsoft&hellip;","msr_research_lab":[199565],"related-researchers":[{"type":"user_nicename","display_name":"Lester Mackey","user_id":36161,"people_section":"Group name 1","alias":"lmackey"},{"type":"user_nicename","display_name":"Ryota Tomioka","user_id":33483,"people_section":"Group name 1","alias":"ryoto"}],"msr_impact_theme":[],"related-academic-programs":[],"related-groups":[144931],"related-projects":[416408],"related-opportunities":[],"related-publications":[394229,915996],"related-videos":[425034,425094,427524,427566],"related-posts":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/394283","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":12,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/394283\/revisions"}],"predecessor-version":[{"id":1147161,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/394283\/revisions\/1147161"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=394283"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=394283"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=394283"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=394283"},{"taxonomy":"msr-video-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-video-type?post=394283"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=394283"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=394283"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=394283"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=394283"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}