{"id":425610,"date":"2017-10-04T13:40:05","date_gmt":"2017-10-04T20:40:05","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-event&p=425610"},"modified":"2025-08-06T11:57:39","modified_gmt":"2025-08-06T18:57:39","slug":"microsoft-research-nips-2017","status":"publish","type":"msr-event","link":"https:\/\/www.microsoft.com\/en-us\/research\/event\/microsoft-research-nips-2017\/","title":{"rendered":"Microsoft @ NIPS 2017"},"content":{"rendered":"\n\n

Venue:\u00a0<\/strong>Long Beach Convention Center (opens in new tab)<\/span><\/a><\/p>\n

Symposia:<\/strong> December 4, 2017<\/p>\n

Demonstrations:<\/strong> December 5\u20136, 2017<\/p>\n

Workshops:<\/strong> December 8\u20139, 2017<\/p>\n

Website:<\/strong>\u00a0Neural Information Processing Systems (opens in new tab)<\/span><\/a><\/p>\n

Machine learning opportunities at Microsoft:<\/strong><\/p>\n

Click here for opportunities in China (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in India (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in Munich (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in the UK (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in the US (opens in new tab)<\/span><\/a><\/p>\n

Paper Legend card game:<\/strong><\/p>\n

Printable version – part 1 (opens in new tab)<\/span><\/a>, part 2 (opens in new tab)<\/span><\/a><\/p>\n

Cloud AI Research Challenge:<\/strong><\/p>\n

\"Microsoft (opens in new tab)<\/span><\/a>Opens in a new tab<\/span><\/p>\n

Microsoft is excited to be a Platinum sponsor of the thirty-first annual conference on Neural Information Processing Systems (NIPS) (opens in new tab)<\/span><\/a>. Over 80 of our researchers are involved in spotlight sessions, presentations, symposiums, posters, accepted papers, and workshops at NIPS (see schedule below). Stop by our booth (#315, Exhibit Hall B) to see HoloLens<\/a> and Windows Mixed Reality<\/a> in action, as well as to find out about career opportunities at Microsoft and enter for your chance to win an Xbox One X (opens in new tab)<\/span><\/a> gaming console. Follow @MSFTResearch (opens in new tab)<\/span><\/a> for the latest information coming from the event.<\/p>\n

NIPS organizing committee<\/h2>\n

Hanna Wallach<\/a>, Program Co-chair
\nJenn Wortman Vaughan<\/a>, Tutorial chair
\nMarkus Weimer<\/a>,\u00a0Demonstration and competition chair<\/p>\n

Workshop organizers<\/h2>\n

Evelyne Viegas<\/a>, Machine Learning Challenges as a Research Tool (opens in new tab)<\/span><\/a>
\nNicolo Fusi<\/a>, Machine Learning in Computational Biology (opens in new tab)<\/span><\/a>
\nSiddhartha Sen<\/a>, ML Systems Workshop (opens in new tab)<\/span><\/a>
\nAlekh Agarwal<\/a>, OPT 2017: Optimization for Machine Learning (opens in new tab)<\/span><\/a>
\nJennifer Wortman Vaughan<\/a>, Learning in the Presence of Strategic Behavior (opens in new tab)<\/span><\/a>
\nManik Varma<\/a>, Extreme Classification: Multi-class & Multi-label Learning in Extremely Large Label Spaces (opens in new tab)<\/span><\/a>
\nVasilis Syrgkanis<\/a>, Learning in the Presence of Strategic Behavior (opens in new tab)<\/span><\/a><\/p>\n

Symposium\u00a0organizers<\/h2>\n

Patrice Simard<\/a>, Interpretable Machine Learning (opens in new tab)<\/span><\/a>
\nRich Caruana<\/a>, Interpretable Machine Learning (opens in new tab)<\/span><\/a><\/p>\n

Invited speaker<\/h2>\n

December 5 @ 1:50\u20132:40 PM | Kate Crawford<\/span><\/a>, <\/span>The Trouble with Bias<\/span> (opens in new tab)<\/span><\/a><\/p>\n

Careers at Microsoft information session<\/h2>\n

December 6 @ 12:30\u20131:20 PM |\u00a0Eric Horvitz<\/a>, Christopher Bishop<\/a>, Jennifer Chayes<\/a>, and Mir Rosenberg<\/p>\n

Spotlight sessions<\/h2>\n

December 5 @ 3:30\u20133:35 PM | Clustering Billions of Reads for DNA Data Storage (opens in new tab)<\/span><\/a>
\nCyrus Rashtchian, Konstantin Makarychev, Luis Ceze, Karin Strauss<\/a>, Sergey Yekhanin<\/a>, Djordje Jevdjic, Miklos Racz, and Siena Ang<\/p>\n

December 6 @ 11:20\u201311:25 AM | Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues
\n<\/a>Noga Alon, Moshe Babaioff<\/a>, Yannai A. Gonczarowski, Yishay Mansour, Shay Moran, and Amir Yehudayoff<\/p>\n

December 6 @ 5:15\u20135:20 PM | Repeated Inverse Reinforcement Learning (opens in new tab)<\/span><\/a>
\nKareem Amin, Nan Jiang<\/a>, and Satinder Singh<\/p>\n

Oral presentations<\/h2>\n

December 5 @ 10:55\u201311:00 AM | Robust Optimization for Non-Convex Objectives
\n<\/a>Yaron Singer, Robert S Chen, Vasilis Syrgkanis<\/a>, and\u00a0Brendan Lucier<\/a><\/p>\n

December 6 @ 4:20\u20134:35 PM | Off-policy evaluation for slate recommendation (opens in new tab)<\/span><\/a>
\nAdith Swaminathan<\/a>, Akshay Krishnamurthy, Alekh Agarwal<\/a>, Miro Dudik<\/a>, John Langford<\/a>, Damien Jose, and\u00a0Imed Zitouni<\/a>Opens in a new tab<\/span><\/p>\n

“A Decomposition of Forecast Error in Prediction Markets<\/a>” by\u00a0Miro Dudik<\/a>, Sebastien Lahaie, Ryan M Rogers, and\u00a0Jennifer Wortman Vaughan<\/a><\/p>\n

“A Highly Efficient Gradient Boosting Decision Tree<\/a>” by\u00a0Guolin Ke, Qi Meng, Taifeng Wang<\/a>, Wei Chen<\/a>, Weidong Ma<\/a>, and\u00a0Tie-Yan Liu<\/a><\/p>\n

“A Sample Complexity Measure with Applications to Learning Optimal Auctions<\/a>” by\u00a0Vasilis Syrgkanis<\/a><\/p>\n

“Accuracy First: Selecting a Differential Privacy Level for Accuracy Constrained ERM<\/a>” by\u00a0Steven Wu<\/a>, Bo Waggoner, Seth Neel, Aaron Roth, and Katrina Ligett<\/p>\n

“Adversarial Ranking for Language Generation<\/a>” by\u00a0Dianqi Li, Kevin Lin, Xiaodong He<\/a>, Ming-ting Sun, and\u00a0Zhengyou Zhang<\/a><\/p>\n

“Clustering Billions of Reads for DNA Data Storage<\/a>” by\u00a0Cyrus Rashtchian, Konstantin Makarychev, Luis Ceze, Karin Strauss<\/a>, Sergey Yekhanin<\/a>, Djordje Jevdjic, Miklos Racz, and Siena Ang<\/p>\n

“Collecting Telemetry Data Privately<\/a>” by\u00a0Bolin Ding<\/a>, Janardhan Kulkarni, and\u00a0Sergey Yekhanin<\/a><\/p>\n

“Consistent Robust Regression<\/a>” by\u00a0Kush Bhatia, Prateek Jain<\/a>, and Purushottam Kar<\/p>\n

“Decoding with Value Networks for Neural Machine Translation<\/a>” by\u00a0Di He<\/a>, Hanqing Lu, Yingce Xia, Tao Qin<\/a>, Liwei Wang, and Tieyan Liu<\/p>\n

“Deliberation Networks: Sequence Generation Beyond One-Pass Decoding<\/a>” by\u00a0Yingce Xia, Lijun Wu, Jianxin Lin, Fei Tian<\/a>, Tao Qin<\/a>, and\u00a0Tie-Yan Liu<\/a><\/p>\n

“Efficiency Guarantees from Data<\/a>” by\u00a0Darrell Hoy, Tremor Technologies; Denis Nekipelov, University of Virginia; and\u00a0Vasilis Syrgkanis<\/a>, Microsoft Research<\/p>\n

“Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach<\/a>” by\u00a0Emmanouil Platanios, Carnegie Mellon University;\u00a0Hoifung Poon<\/a>, Microsoft Research; Tom M. Mitchell, Carnegie Mellon University; and\u00a0Eric J. Horvitz<\/a>, Microsoft Research<\/p>\n

“From Bayesian Sparsity to Gated Recurrent Nets<\/a>” by\u00a0Hao He, Massachusetts Institute of Technology;\u00a0Bo Xin<\/a>, Microsoft Research; and\u00a0David Wipf<\/a>, Microsoft Research<\/p>\n

“Hybrid Reward Architecture for Reinforcement Learning<\/a>” by\u00a0Harm Van Seijen<\/a>, Microsoft Research;\u00a0Romain Laroche<\/a>, Microsoft Research, Maluuba;\u00a0Mehdi Fatemi<\/a>, Microsoft Research; and Joshua Romoff, McGill University<\/p>\n

“Identifying Outlier Arms in Multi-Armed Bandit<\/a>” by\u00a0Honglei Zhuang, University of Illinois;\u00a0Chi Wang<\/a>, Microsoft Research; and Yifan Wang, Tsinghua University<\/p>\n

“Improving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications<\/a>” by\u00a0Qinshi Wang and\u00a0Wei Chen<\/a><\/p>\n

“Inference in Graphical Models via Semidefinite Programming Hierarchies<\/a>” by\u00a0Murat Erdogdu<\/a>, Yash Deshpande, and Andrea Montanari<\/p>\n

“Influence Maximization with \u03b5<\/span><\/span><\/span><\/span>-Almost Submodular Threshold Function<\/a>” by Qiang Li, Institute of Computing Technol;\u00a0Wei Chen<\/a>, Microsoft Research; Xiaoming Sun, Institute of Computing Technology, Chinese Academy of Sciences; and Jialin Zhang, Institute of Computing Technology, Chinese Academy of Sciences<\/p>\n

“Large-Scale Quadratically Constrained Quadratic Program via Low-Discrepancy Sequences<\/a>” by\u00a0Kinjal Basu,\u00a0Ankan Saha, and\u00a0Shaunak Chatterjee,\u00a0LinkedIn Corporation<\/p>\n

“Learning Mixture of Gaussians with Streaming Data<\/a>” by\u00a0Aditi Raghunathan, Stanford University;\u00a0Prateek Jain<\/a>, Microsoft Research; and Ravishankar Krishnawamy, Microsoft Research<\/p>\n

“Linear Convergence of a Frank-Wolfe Type Algorithm over Trace-Norm Balls<\/a>” by\u00a0Zeyuan Allen-Zhu<\/a>, Microsoft Research; Elad Hazan, Princeton University; Wei Hu, Princeton University; Yuanzhi Li, Princeton University<\/p>\n

“The Importance of Communities for Learning to Influence<\/a>” by\u00a0Eric Balkanski, Harvard University;\u00a0Nicole Immorlica<\/a>, Microsoft Research; and Yaron Singer, Harvard University<\/p>\n

“Mean Field Residual Networks: On the Edge of Chaos<\/a>” by Greg Yang, Microsoft Research; Samuel S. Schoenholz, Google Brain<\/p>\n

“Multi-Task Learning for Contextual Bandits<\/a>” by\u00a0Aniket Anand Deshmukh, University of Michigan, Ann Arbor; Urun Dogan, Microsoft; and Clay Scott, University of Michigan<\/p>\n

“Neural Program Meta-Induction<\/a>” by\u00a0Jacob Devlin, Microsoft Research; Rudy R Bunel, Oxford University;\u00a0Rishabh Singh<\/a>, Microsoft Research;\u00a0Matthew Hausknecht<\/a>, Microsoft Research; and Pushmeet Kohli, DeepMind<\/p>\n

“Non-convex Robust PCA<\/a>” by Praneeth Netrapalli, Microsoft Research; Niranjan Uma Naresh, UC Irvine; Sujay Sanghavi, UT-Austin; Animashree Anadkumar, UC-Irvine; Prateek Jain, Microsoft Research<\/p>\n

“Off-policy Evaluation for Slate Recommendation<\/a>” by\u00a0Adith Swaminathan<\/a>, Microsoft Research; Akshay Krishnamurthy, University of Massachusetts;\u00a0Alekh Agarwal<\/a>, Microsoft Research;\u00a0Miro Dudik<\/a>, Microsoft Research;\u00a0John Langford<\/a>, Microsoft Research; Damien Jose, Microsoft; and\u00a0Imed Zitouni<\/a>, Microsoft Research<\/p>\n

“Online Learning with a Hint<\/a>” by\u00a0Ofer Dekel<\/a>, Microsoft Research; Arthur Flajolet, Massachusetts Institute of Technology; Nika Haghtalab, Carnegie Mellon University; and Patrick Jaillet, Massachusetts Institute of Technology<\/p>\n

“Plan, Attend, Generate: Planning for Sequence-to-Sequence Models<\/a>” by\u00a0Caglar Gulcehre, Deepmind; Francis Dutil, MILA; Adam Trischler, Microsoft; and Yoshua Bengio, University of Montreal<\/p>\n

“Q-LDA: Uncovering Latent Patterns in Text-based Sequential Decision Processes<\/a>” by\u00a0Jianshu Chen<\/a>, Microsoft Research; Chong Wang, Princeton University;\u00a0Lin Xiao<\/a>, Microsoft Research; Ji He, University Washington;\u00a0Lihong Li<\/a>, Microsoft Research; and Li Deng, Citadel<\/p>\n

“QSGD: Communication-Efficient Stochastic Gradient Descent, with Applications to Neural Networks<\/a>” by\u00a0Dan Alistarh, Demjan Grubic, Jerry Li, Ryota Tomioka<\/a>, and Milan Vojnovic<\/p>\n

“Repeated Inverse Reinforcement Learning<\/a>” by\u00a0Kareem Amin, Google Research;\u00a0Nan Jiang<\/a>, Microsoft Research; and Satinder Singh, University of Michigan<\/p>\n

“Robust Estimation of Neural Signals in Calcium Imaging<\/a>” by\u00a0Hakan Inan, Stanford University;\u00a0Murat Erdogdu<\/a>, Microsoft Research; and Mark Schnitzer, Stanford University<\/p>\n

“Robust Optimization for Non-Convex Objectives<\/a>” by\u00a0Yaron Singer, Harvard University; Robert S Chen, Harvard University;\u00a0Vasilis Syrgkanis<\/a>, Microsoft Research; and\u00a0Brendan Lucier<\/a>, Microsoft Research<\/p>\n

“Stabilizing Training of Generative Adversarial Networks through Regularization<\/a>” by\u00a0Kevin Roth, ETH; Aurelien Lucchi, ETH Zurich;\u00a0Sebastian Nowozin<\/a>, Microsoft Research; and Thomas Hofmann, ETH Zurich<\/p>\n

“Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues<\/a>” by\u00a0Noga Alon, Tel Aviv University;\u00a0Moshe Babaioff<\/a>, Microsoft Research; Yannai A. Gonczarowski, The Hebrew University of Jerusalem and Microsoft Research; Yishay Mansour, Tel Aviv University; Shay Moran, IAS, Princeton; and Amir Yehudayoff, Technion – Israel Institute of Technology<\/p>\n

“The Numerics of GANs<\/a>” by\u00a0Lars Mescheder, Max-Planck Institute Tuebingen;\u00a0Sebastian Nowozin<\/a>, Microsoft Research; and Andreas Geiger, MPI T\u00fcbingen<\/p>\n

“Thy Friend is My Friend: Iterative Collaborative Filtering for Sparse Matrix Estimation<\/a>” by\u00a0Christian Borgs<\/a>, Microsoft Research;\u00a0Jennifer Chayes<\/a>, Microsoft Research;\u00a0Christina Lee<\/a>, Microsoft Research; and Devavrat Shah, Massachusetts Institute of Technology<\/p>\n

“Unsupervised Sequence Classification using Sequential Output Statistics<\/a>” by\u00a0Yu Liu, SUNY Buffalo;\u00a0Jianshu Chen<\/a>, Microsoft Research; and Li Deng, Citadel<\/p>\n

“Z-Forcing: Training Stochastic Recurrent Networks<\/a>” by\u00a0Marc-Alexandre C\u00f4t\u00e9, Microsoft Research; Alessandro Sordoni, Microsoft Research, Maluuba; Anirudh Goyal, Universit\u00e9 de Montr\u00e9al; Nan Ke, MILA, \u00c9cole Polytechnique de Montr\u00e9al; and Yoshua Bengio, University of MontrealOpens in a new tab<\/span><\/p>\n

A Decomposition of Forecast Error in Prediction Markets<\/a>
\nMiro Dudik<\/a> (Microsoft Research),\u00a0Jennifer Wortman Vaughan<\/a>\u00a0(Microsoft Research)<\/p>\n

A Highly Efficient Gradient Boosting Decision Tree<\/a>
\nGuolin Ke (Microsoft Research), Taifeng Wang<\/a> (Microsoft Research), Wei Chen<\/a> (Microsoft Research), Weidong Ma<\/a> (Microsoft Research), Tie-Yan Liu<\/a>\u00a0(Microsoft Research)<\/p>\n

A Sample Complexity Measure with Applications to Learning Optimal Auctions<\/a>
\nVasilis Syrgkanis<\/a>\u00a0(Microsoft Research)<\/p>\n

Adversarial Ranking for Language Generation<\/a>
\nXiaodong He<\/a> (Microsoft Research), Zhengyou Zhang<\/a>\u00a0(Microsoft Research)<\/p>\n

Clustering Billions of Reads for DNA Data Storage<\/a>
\nLuis Ceze,\u00a0Karin Strauss<\/a> (Microsoft Research),\u00a0Sergey Yekhanin<\/a> (Microsoft Research),\u00a0Djordje Jevdjic (Microsoft Research),\u00a0Siena Ang (Microsoft), Konstantin Makarychev (Microsoft)<\/p>\n

Collecting Telemetry Data Privately<\/a>
\nBolin Ding<\/a> (Microsoft Research),\u00a0Janardhan Kulkarni (Microsoft Research),\u00a0Sergey Yekhanin<\/a>\u00a0(Microsoft Research)<\/p>\n

Communication-Efficient Stochastic Gradient Descent, with Applications to Neural Networks<\/a>
\nRyota Tomioka<\/a>\u00a0(Microsoft Research)<\/p>\n

Consistent Robust Regression<\/a>
\nPrateek Jain<\/a>\u00a0(Microsoft Research)<\/p>\n

Decoding with Value Networks for Neural Machine Translation<\/a>
\nDi He<\/a>, Tao Qin<\/a> (Microsoft Research), Tieyan Liu (Microsoft Research)<\/p>\n

Deliberation Networks: Sequence Generation Beyond One-Pass Decoding<\/a>
\nJianxin Lin, Fei Tian<\/a> (Microsoft Research), Tao Qin<\/a> (Microsoft Research), Tie-Yan Liu<\/a>\u00a0(Microsoft Research)<\/p>\n

Efficiency Guarantees from Data<\/a>
\nVasilis Syrgkanis<\/a>\u00a0(Microsoft Research)<\/p>\n

Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach<\/a>
\nHoifung Poon<\/a> (Microsoft Research), Eric Horvitz<\/a>\u00a0(Microsoft Research)<\/p>\n

From Bayesian Sparsity to Gated Recurrent Nets<\/a>
\nDavid Wipf<\/a>\u00a0(Microsoft Research)<\/p>\n

Hybrid Reward Architecture for Reinforcement Learning<\/a>
\nHarm Van Seijen<\/a> (Microsoft Research), Mehdi Fatemi<\/a>\u00a0(Microsoft Research)<\/p>\n

Identifying Outlier Arms in Multi-Armed Bandit<\/a>
\nChi Wang<\/a>\u00a0(Microsoft Research)<\/p>\n

Improving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications<\/a>
\nWei Chen<\/a>\u00a0(Microsoft Research)<\/p>\n

Influence Maximization with \u03b5<\/span><\/span><\/span><\/span>-Almost Submodular Threshold Function<\/a>
\nWei Chen<\/a>\u00a0(Microsoft Research)<\/p>\n

Large-Scale Quadratically Constrained Quadratic Program via Low-Discrepancy Sequences<\/a>
\nKinjal Basu (LinkedIn Corporation),\u00a0Ankan Saha (LinkedIn Corporation), Shaunak Chatterjee (LinkedIn Corporation)<\/p>\n

Learning Mixture of Gaussians with Streaming Data<\/a>
\nPrateek Jain<\/a>\u00a0(Microsoft Research)<\/p>\n

Neural Program Meta-Induction<\/a>
\nJacob Devlin (Microsoft Research), Rishabh Singh<\/a>\u00a0(Microsoft Research), Matthew Hausknecht<\/a>\u00a0(Microsoft Research)<\/p>\n

Off-policy Evaluation for Slate Recommendation<\/a>
\nAdith Swaminathan<\/a> (Microsoft Research),\u00a0Alekh Agarwal<\/a> (Microsoft Research),\u00a0Miro Dudik<\/a> (Microsoft Research),\u00a0Damien Jose (Microsoft),\u00a0Imed Zitouni<\/a>\u00a0(Microsoft Research)<\/p>\n

Online Learning with a Hint<\/a>
\nOfer Dekel<\/a>\u00a0(Microsoft Research)<\/p>\n

Plan, Attend, Generate: Planning for Sequence-to-Sequence Models<\/a>
\nAdam Trischler (Microsoft)<\/p>\n

Q-LDA: Uncovering Latent Patterns in Text-based Sequential Decision Processes<\/a>
\nJianshu Chen<\/a> (Microsoft Research), Lin Xiao<\/a>\u00a0(Microsoft Research)<\/p>\n

Robust Optimization for Non-Convex Objectives<\/a>
\nVasilis Syrgkanis<\/a> (Microsoft Research), Brendan Lucier<\/a>\u00a0(Microsoft Research)<\/p>\n

Stabilizing Training of Generative Adversarial Networks through Regularization<\/a>
\nSebastian Nowozin<\/a> (Microsoft Research)<\/p>\n

Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues<\/a>
\nMoshe Babaioff<\/a>\u00a0(Microsoft Research)<\/p>\n

The Numerics of GANs<\/a>
\nSebastian Nowozin<\/a>\u00a0(Microsoft Research)<\/p>\n

Thy Friend is My Friend: Iterative Collaborative Filtering for Sparse Matrix Estimation<\/a>
\nChristian Borgs<\/a> (Microsoft Research),\u00a0Jennifer Chayes<\/a>\u00a0(Microsoft Research)<\/p>\n

Unsupervised Sequence Classification using Sequential Output Statistics<\/a>
\nJianshu Chen<\/a>\u00a0(Microsoft Research)<\/p>\n

Z-Forcing: Training Stochastic Recurrent Networks<\/a>
\nMarc-Alexandre C\u00f4t\u00e9 (Microsoft Research),\u00a0Alessandro Sordoni\u00a0(Microsoft Research, Maluuba)Opens in a new tab<\/span><\/p>\n

NIPS 2017 Workshops<\/h2>\n

From \u201cWhat if?\u201d To \u201cWhat Next?\u201d: Causal Inference and Machine Learning for Intelligent Decision Making<\/h3>\n

Friday, December 8 @ 9:00 AM\u20136:30 PM | Hall C | Adith Swaminathan<\/a>, Microsoft Research<\/p>\n

This workshop is aimed at facilitating more interactions between researchers in machine learning and causal inference. In particular, it is an opportunity to bring together highly technical individuals who are strongly motivated by the practical importance and real-world impact of their work. Cultivating such interactions will lead to the development of theory, methodology, and – most importantly – practical tools, that better target causal questions across different domains.<\/p>\n

In particular, we will highlight theory, algorithms and applications on automatic decision making systems, such as recommendation engines, medical decision systems and self-driving cars, as both producers and users of data. The challenge here is the feedback between learning from data and then taking actions that may affect what data will be made available for future learning. Learning algorithms have to reason about how changes to the system will affect future data, giving rise to challenging counterfactual and causal reasoning issues that the learning algorithm has to account for. Modern and scalable policy learning algorithms also require operating with non-experimental data, such as logged user interaction data where users click ads suggested by recommender systems trained on historical user clicks.<\/p>\n

To further bring the community together around the use of such interaction data, this workshop will host a Kaggle challenge problem based on the first real-world dataset of logged contextual bandit feedback with non-uniform action-selection propensities. The dataset consists of several gigabytes of data from an ad placement system, which we have processed into multiple well-defined learning problems of increasing complexity, feedback signal, and context. Participants in the challenge problem will be able to discuss their results at the workshop.<\/p>\n

Machine Learning and Computer Security<\/h3>\n

Friday, December 8 @ 9:00 AM\u20135:00 PM | Hyatt Hotel, Shoreline | Donald Brinkman<\/a>, Microsoft Research<\/p>\n

While traditional computer security relies on well-defined attack models and proofs of security, a science of security for machine learning systems has proven more elusive. This is due to a number of obstacles, including (1) the highly varied angles of attack against ML systems, (2) the lack of a clearly defined attack surface (because the source of the data analyzed by ML systems is not easily traced), and (3) the lack of clear formal definitions of security that are appropriate for ML systems. At the same time, security of ML systems is of great import due the recent trend of using ML systems as a line of defense against malicious behavior (e.g., network intrusion, malware, and ransomware), as well as the prevalence of ML systems as parts of sensitive and valuable software systems (e.g., sentiment analyzers for predicting stock prices). This workshop will bring together experts from the computer security and machine learning communities in an attempt to highlight recent work in this area, as well as to clarify the foundations of secure ML and chart out important directions for future work and cross-community collaborations.<\/p>\n

Conversational AI – today’s practice and tomorrow’s potential<\/h3>\n

Friday, December 8 @ 8:00 AM\u20137:00 PM | Grand Ballroom B | Jason Williams<\/a>, Microsoft Research<\/p>\n

This workshop will include invited talks from academia and industry, contributed work, and open discussion. In these talks, senior technical leaders from many of the most popular conversational services will give insights into real usage and challenges at scale. An open call for papers will be issued, and we will prioritize forward-looking papers that propose interesting and impactful contributions. We will end the day with an open discussion, including a panel consisting of academic and industrial researchers.<\/p>\n

Interpreting, Explaining and Visualizing Deep Learning\u2026now what?<\/h3>\n

Saturday, December 9 @ 8:15 AM\u20136:30 PM | Hyatt Regency Ballroom | Hamid Palangi<\/a>, Qiuyuan Huang<\/a>, Paul Smolensky<\/a>, and Xiaodong He<\/a>, Microsoft Research<\/p>\n

Our NIPS 2017 Workshop \u201cInterpreting, Explaining and Visualizing Deep Learning \u2013 Now what?\u201d aims to review recent techniques and establish new theoretical foundations for interpreting and understanding deep learning models. However, it will not stop at the methodological level, but also address the \u201cnow what?\u201d question. This strong focus on the applications of interpretable methods in deep learning distinguishes this workshop from previous events as we aim to take the next step by exploring and extending the practical usefulness of Interpreting, Explaining and Visualizing in Deep Learning. Also with this workshop we aim to identify new fields of applications for interpretable deep learning. Since the workshop will host invited speakers from various application domains (computer vision, NLP, neuroscience, medicine), it will provide an opportunity for participants to learn from each other and initiate new interdisciplinary collaborations. The workshop will contain invited research talks, short methods and applications talks, a poster and demonstration session and a panel discussion. A selection of accepted papers together with the invited contributions will be published in an edited book by Springer LNCS in order to provide a representative overview of recent activities in this emerging research field.<\/p>\n

Co-located workshops<\/h2>\n

Women in Machine Learning<\/h3>\n

Monday, December 4 & Thursday, December 7 @ 2:00 PM\u20132:30 PM | Room 104 | 12th Women in Machine Learning Workshop (WiML 2017) (opens in new tab)<\/span><\/a>, by Hanna Wallach<\/a>, Microsoft Research<\/p>\n

The annual Women in Machine Learning Workshop is the\u00a0flagship event of Women in Machine Learning (opens in new tab)<\/span><\/a>.\u00a0This technical workshop gives female faculty, research scientists, and graduate students in the machine learning community an opportunity to meet, network and exchange ideas, participate in career-focused panel discussions with senior women in industry and academia and learn from each other. Underrepresented minorities and undergraduates interested in machine learning research are encouraged to attend.\u00a0We welcome all genders; however, any formal presentations, i.e. talks and posters, are given\u00a0by women.\u00a0We strive to create an atmosphere in which participants feel comfortable to engage in\u00a0technical and career-related conversations.<\/p>\n

Black in AI<\/h3>\n

Friday, December 8 @ 1:30 PM\u20135:30 PM | Black in AI Workshop @ NIPS 2017<\/a>, by Timnit Gebru<\/a>, Microsoft Research<\/p>\n

The first Black in AI event will be co-located with NIPS 2017. The goal is to gather people in the field to share ideas and discuss\u00a0initiatives to increase the presence of Black\u00a0people in the field of\u00a0artificial intelligence, for both diversity and data bias prevention purposes. At this workshop, Black researchers in AI will also have the opportunity to present their work during our oral and poster sessions.Opens in a new tab<\/span><\/p>\n","protected":false},"excerpt":{"rendered":"

The Thirty-first Annual Conference on Neural Information Processing Systems (NIPS) is a multi-track machine learning and computational neuroscience conference that includes invited talks, demonstrations, symposia and oral and poster presentations of refereed papers.<\/p>\n","protected":false},"featured_media":446514,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_startdate":"2017-12-04","msr_enddate":"2017-12-09","msr_location":"Long Beach, California","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-425610","msr-event","type-msr-event","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-region-north-america","msr-event-type-conferences","msr-locale-en_us"],"msr_about":"\n\n

Venue:\u00a0<\/strong>Long Beach Convention Center (opens in new tab)<\/span><\/a><\/p>\n

Symposia:<\/strong> December 4, 2017<\/p>\n

Demonstrations:<\/strong> December 5\u20136, 2017<\/p>\n

Workshops:<\/strong> December 8\u20139, 2017<\/p>\n

Website:<\/strong>\u00a0Neural Information Processing Systems (opens in new tab)<\/span><\/a><\/p>\n

Machine learning opportunities at Microsoft:<\/strong><\/p>\n

Click here for opportunities in China (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in India (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in Munich (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in the UK (opens in new tab)<\/span><\/a><\/p>\n

Click here for opportunities in the US (opens in new tab)<\/span><\/a><\/p>\n

Paper Legend card game:<\/strong><\/p>\n

Printable version – part 1 (opens in new tab)<\/span><\/a>, part 2 (opens in new tab)<\/span><\/a><\/p>\n

Cloud AI Research Challenge:<\/strong><\/p>\n

\"Microsoft (opens in new tab)<\/span><\/a>Opens in a new tab<\/span><\/p>\n

Microsoft is excited to be a Platinum sponsor of the thirty-first annual conference on Neural Information Processing Systems (NIPS)<\/a>. Over 80 of our researchers are involved in spotlight sessions, presentations, symposiums, posters, accepted papers, and workshops at NIPS (see schedule below). Stop by our booth (#315, Exhibit Hall B) to see HoloLens<\/a> and Windows Mixed Reality<\/a> in action, as well as to find out about career opportunities at Microsoft and enter for your chance to win an Xbox One X<\/a> gaming console. Follow @MSFTResearch<\/a> for the latest information coming from the event.<\/p>\n

NIPS organizing committee<\/h2>\n

Hanna Wallach<\/a>, Program Co-chair
\nJenn Wortman Vaughan<\/a>, Tutorial chair
\nMarkus Weimer<\/a>,\u00a0Demonstration and competition chair<\/p>\n

Workshop organizers<\/h2>\n

Evelyne Viegas<\/a>, Machine Learning Challenges as a Research Tool<\/a>
\nNicolo Fusi<\/a>, Machine Learning in Computational Biology<\/a>
\nSiddhartha Sen<\/a>, ML Systems Workshop<\/a>
\nAlekh Agarwal<\/a>, OPT 2017: Optimization for Machine Learning<\/a>
\nJennifer Wortman Vaughan<\/a>, Learning in the Presence of Strategic Behavior<\/a>
\nManik Varma<\/a>, Extreme Classification: Multi-class & Multi-label Learning in Extremely Large Label Spaces<\/a>
\nVasilis Syrgkanis<\/a>, Learning in the Presence of Strategic Behavior<\/a><\/p>\n

Symposium\u00a0organizers<\/h2>\n

Patrice Simard<\/a>, Interpretable Machine Learning<\/a>
\nRich Caruana<\/a>, Interpretable Machine Learning<\/a><\/p>\n

Invited speaker<\/h2>\n

December 5 @ 1:50\u20132:40 PM | Kate Crawford<\/span><\/a>, <\/span>The Trouble with Bias<\/span><\/a><\/p>\n

Careers at Microsoft information session<\/h2>\n

December 6 @ 12:30\u20131:20 PM |\u00a0Eric Horvitz<\/a>, Christopher Bishop<\/a>, Jennifer Chayes<\/a>, and Mir Rosenberg<\/p>\n

Spotlight sessions<\/h2>\n

December 5 @ 3:30\u20133:35 PM | Clustering Billions of Reads for DNA Data Storage<\/a>
\nCyrus Rashtchian, Konstantin Makarychev, Luis Ceze, Karin Strauss<\/a>, Sergey Yekhanin<\/a>, Djordje Jevdjic, Miklos Racz, and Siena Ang<\/p>\n

December 6 @ 11:20\u201311:25 AM | Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues
\n<\/a>Noga Alon, Moshe Babaioff<\/a>, Yannai A. Gonczarowski, Yishay Mansour, Shay Moran, and Amir Yehudayoff<\/p>\n

December 6 @ 5:15\u20135:20 PM | Repeated Inverse Reinforcement Learning<\/a>
\nKareem Amin, Nan Jiang<\/a>, and Satinder Singh<\/p>\n

Oral presentations<\/h2>\n

December 5 @ 10:55\u201311:00 AM | Robust Optimization for Non-Convex Objectives
\n<\/a>Yaron Singer, Robert S Chen, Vasilis Syrgkanis<\/a>, and\u00a0Brendan Lucier<\/a><\/p>\n

December 6 @ 4:20\u20134:35 PM | Off-policy evaluation for slate recommendation<\/a>
\nAdith Swaminathan<\/a>, Akshay Krishnamurthy, Alekh Agarwal<\/a>, Miro Dudik<\/a>, John Langford<\/a>, Damien Jose, and\u00a0Imed Zitouni<\/a>Opens in a new tab<\/span><\/p>\n

“A Decomposition of Forecast Error in Prediction Markets<\/a>” by\u00a0Miro Dudik<\/a>, Sebastien Lahaie, Ryan M Rogers, and\u00a0Jennifer Wortman Vaughan<\/a><\/p>\n

“A Highly Efficient Gradient Boosting Decision Tree<\/a>” by\u00a0Guolin Ke, Qi Meng, Taifeng Wang<\/a>, Wei Chen<\/a>, Weidong Ma<\/a>, and\u00a0Tie-Yan Liu<\/a><\/p>\n

“A Sample Complexity Measure with Applications to Learning Optimal Auctions<\/a>” by\u00a0Vasilis Syrgkanis<\/a><\/p>\n

“Accuracy First: Selecting a Differential Privacy Level for Accuracy Constrained ERM<\/a>” by\u00a0Steven Wu<\/a>, Bo Waggoner, Seth Neel, Aaron Roth, and Katrina Ligett<\/p>\n

“Adversarial Ranking for Language Generation<\/a>” by\u00a0Dianqi Li, Kevin Lin, Xiaodong He<\/a>, Ming-ting Sun, and\u00a0Zhengyou Zhang<\/a><\/p>\n

“Clustering Billions of Reads for DNA Data Storage<\/a>” by\u00a0Cyrus Rashtchian, Konstantin Makarychev, Luis Ceze, Karin Strauss<\/a>, Sergey Yekhanin<\/a>, Djordje Jevdjic, Miklos Racz, and Siena Ang<\/p>\n

“Collecting Telemetry Data Privately<\/a>” by\u00a0Bolin Ding<\/a>, Janardhan Kulkarni, and\u00a0Sergey Yekhanin<\/a><\/p>\n

“Consistent Robust Regression<\/a>” by\u00a0Kush Bhatia, Prateek Jain<\/a>, and Purushottam Kar<\/p>\n

“Decoding with Value Networks for Neural Machine Translation<\/a>” by\u00a0Di He<\/a>, Hanqing Lu, Yingce Xia, Tao Qin<\/a>, Liwei Wang, and Tieyan Liu<\/p>\n

“Deliberation Networks: Sequence Generation Beyond One-Pass Decoding<\/a>” by\u00a0Yingce Xia, Lijun Wu, Jianxin Lin, Fei Tian<\/a>, Tao Qin<\/a>, and\u00a0Tie-Yan Liu<\/a><\/p>\n

“Efficiency Guarantees from Data<\/a>” by\u00a0Darrell Hoy, Tremor Technologies; Denis Nekipelov, University of Virginia; and\u00a0Vasilis Syrgkanis<\/a>, Microsoft Research<\/p>\n

“Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach<\/a>” by\u00a0Emmanouil Platanios, Carnegie Mellon University;\u00a0Hoifung Poon<\/a>, Microsoft Research; Tom M. Mitchell, Carnegie Mellon University; and\u00a0Eric J. Horvitz<\/a>, Microsoft Research<\/p>\n

“From Bayesian Sparsity to Gated Recurrent Nets<\/a>” by\u00a0Hao He, Massachusetts Institute of Technology;\u00a0Bo Xin<\/a>, Microsoft Research; and\u00a0David Wipf<\/a>, Microsoft Research<\/p>\n

“Hybrid Reward Architecture for Reinforcement Learning<\/a>” by\u00a0Harm Van Seijen<\/a>, Microsoft Research;\u00a0Romain Laroche<\/a>, Microsoft Research, Maluuba;\u00a0Mehdi Fatemi<\/a>, Microsoft Research; and Joshua Romoff, McGill University<\/p>\n

“Identifying Outlier Arms in Multi-Armed Bandit<\/a>” by\u00a0Honglei Zhuang, University of Illinois;\u00a0Chi Wang<\/a>, Microsoft Research; and Yifan Wang, Tsinghua University<\/p>\n

“Improving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications<\/a>” by\u00a0Qinshi Wang and\u00a0Wei Chen<\/a><\/p>\n

“Inference in Graphical Models via Semidefinite Programming Hierarchies<\/a>” by\u00a0Murat Erdogdu<\/a>, Yash Deshpande, and Andrea Montanari<\/p>\n

“Influence Maximization with \u03b5<\/span><\/span><\/span><\/span>-Almost Submodular Threshold Function<\/a>” by Qiang Li, Institute of Computing Technol;\u00a0Wei Chen<\/a>, Microsoft Research; Xiaoming Sun, Institute of Computing Technology, Chinese Academy of Sciences; and Jialin Zhang, Institute of Computing Technology, Chinese Academy of Sciences<\/p>\n

“Large-Scale Quadratically Constrained Quadratic Program via Low-Discrepancy Sequences<\/a>” by\u00a0Kinjal Basu,\u00a0Ankan Saha, and\u00a0Shaunak Chatterjee,\u00a0LinkedIn Corporation<\/p>\n

“Learning Mixture of Gaussians with Streaming Data<\/a>” by\u00a0Aditi Raghunathan, Stanford University;\u00a0Prateek Jain<\/a>, Microsoft Research; and Ravishankar Krishnawamy, Microsoft Research<\/p>\n

“Linear Convergence of a Frank-Wolfe Type Algorithm over Trace-Norm Balls<\/a>” by\u00a0Zeyuan Allen-Zhu<\/a>, Microsoft Research; Elad Hazan, Princeton University; Wei Hu, Princeton University; Yuanzhi Li, Princeton University<\/p>\n

“The Importance of Communities for Learning to Influence<\/a>” by\u00a0Eric Balkanski, Harvard University;\u00a0Nicole Immorlica<\/a>, Microsoft Research; and Yaron Singer, Harvard University<\/p>\n

“Mean Field Residual Networks: On the Edge of Chaos<\/a>” by Greg Yang, Microsoft Research; Samuel S. Schoenholz, Google Brain<\/p>\n

“Multi-Task Learning for Contextual Bandits<\/a>” by\u00a0Aniket Anand Deshmukh, University of Michigan, Ann Arbor; Urun Dogan, Microsoft; and Clay Scott, University of Michigan<\/p>\n

“Neural Program Meta-Induction<\/a>” by\u00a0Jacob Devlin, Microsoft Research; Rudy R Bunel, Oxford University;\u00a0Rishabh Singh<\/a>, Microsoft Research;\u00a0Matthew Hausknecht<\/a>, Microsoft Research; and Pushmeet Kohli, DeepMind<\/p>\n

“Non-convex Robust PCA<\/a>” by Praneeth Netrapalli, Microsoft Research; Niranjan Uma Naresh, UC Irvine; Sujay Sanghavi, UT-Austin; Animashree Anadkumar, UC-Irvine; Prateek Jain, Microsoft Research<\/p>\n

“Off-policy Evaluation for Slate Recommendation<\/a>” by\u00a0Adith Swaminathan<\/a>, Microsoft Research; Akshay Krishnamurthy, University of Massachusetts;\u00a0Alekh Agarwal<\/a>, Microsoft Research;\u00a0Miro Dudik<\/a>, Microsoft Research;\u00a0John Langford<\/a>, Microsoft Research; Damien Jose, Microsoft; and\u00a0Imed Zitouni<\/a>, Microsoft Research<\/p>\n

“Online Learning with a Hint<\/a>” by\u00a0Ofer Dekel<\/a>, Microsoft Research; Arthur Flajolet, Massachusetts Institute of Technology; Nika Haghtalab, Carnegie Mellon University; and Patrick Jaillet, Massachusetts Institute of Technology<\/p>\n

“Plan, Attend, Generate: Planning for Sequence-to-Sequence Models<\/a>” by\u00a0Caglar Gulcehre, Deepmind; Francis Dutil, MILA; Adam Trischler, Microsoft; and Yoshua Bengio, University of Montreal<\/p>\n

“Q-LDA: Uncovering Latent Patterns in Text-based Sequential Decision Processes<\/a>” by\u00a0Jianshu Chen<\/a>, Microsoft Research; Chong Wang, Princeton University;\u00a0Lin Xiao<\/a>, Microsoft Research; Ji He, University Washington;\u00a0Lihong Li<\/a>, Microsoft Research; and Li Deng, Citadel<\/p>\n

“QSGD: Communication-Efficient Stochastic Gradient Descent, with Applications to Neural Networks<\/a>” by\u00a0Dan Alistarh, Demjan Grubic, Jerry Li, Ryota Tomioka<\/a>, and Milan Vojnovic<\/p>\n

“Repeated Inverse Reinforcement Learning<\/a>” by\u00a0Kareem Amin, Google Research;\u00a0Nan Jiang<\/a>, Microsoft Research; and Satinder Singh, University of Michigan<\/p>\n

“Robust Estimation of Neural Signals in Calcium Imaging<\/a>” by\u00a0Hakan Inan, Stanford University;\u00a0Murat Erdogdu<\/a>, Microsoft Research; and Mark Schnitzer, Stanford University<\/p>\n

“Robust Optimization for Non-Convex Objectives<\/a>” by\u00a0Yaron Singer, Harvard University; Robert S Chen, Harvard University;\u00a0Vasilis Syrgkanis<\/a>, Microsoft Research; and\u00a0Brendan Lucier<\/a>, Microsoft Research<\/p>\n

“Stabilizing Training of Generative Adversarial Networks through Regularization<\/a>” by\u00a0Kevin Roth, ETH; Aurelien Lucchi, ETH Zurich;\u00a0Sebastian Nowozin<\/a>, Microsoft Research; and Thomas Hofmann, ETH Zurich<\/p>\n

“Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues<\/a>” by\u00a0Noga Alon, Tel Aviv University;\u00a0Moshe Babaioff<\/a>, Microsoft Research; Yannai A. Gonczarowski, The Hebrew University of Jerusalem and Microsoft Research; Yishay Mansour, Tel Aviv University; Shay Moran, IAS, Princeton; and Amir Yehudayoff, Technion – Israel Institute of Technology<\/p>\n

“The Numerics of GANs<\/a>” by\u00a0Lars Mescheder, Max-Planck Institute Tuebingen;\u00a0Sebastian Nowozin<\/a>, Microsoft Research; and Andreas Geiger, MPI T\u00fcbingen<\/p>\n

“Thy Friend is My Friend: Iterative Collaborative Filtering for Sparse Matrix Estimation<\/a>” by\u00a0Christian Borgs<\/a>, Microsoft Research;\u00a0Jennifer Chayes<\/a>, Microsoft Research;\u00a0Christina Lee<\/a>, Microsoft Research; and Devavrat Shah, Massachusetts Institute of Technology<\/p>\n

“Unsupervised Sequence Classification using Sequential Output Statistics<\/a>” by\u00a0Yu Liu, SUNY Buffalo;\u00a0Jianshu Chen<\/a>, Microsoft Research; and Li Deng, Citadel<\/p>\n

“Z-Forcing: Training Stochastic Recurrent Networks<\/a>” by\u00a0Marc-Alexandre C\u00f4t\u00e9, Microsoft Research; Alessandro Sordoni, Microsoft Research, Maluuba; Anirudh Goyal, Universit\u00e9 de Montr\u00e9al; Nan Ke, MILA, \u00c9cole Polytechnique de Montr\u00e9al; and Yoshua Bengio, University of MontrealOpens in a new tab<\/span><\/p>\n

A Decomposition of Forecast Error in Prediction Markets<\/a>
\nMiro Dudik<\/a> (Microsoft Research),\u00a0Jennifer Wortman Vaughan<\/a>\u00a0(Microsoft Research)<\/p>\n

A Highly Efficient Gradient Boosting Decision Tree<\/a>
\nGuolin Ke (Microsoft Research), Taifeng Wang<\/a> (Microsoft Research), Wei Chen<\/a> (Microsoft Research), Weidong Ma<\/a> (Microsoft Research), Tie-Yan Liu<\/a>\u00a0(Microsoft Research)<\/p>\n

A Sample Complexity Measure with Applications to Learning Optimal Auctions<\/a>
\nVasilis Syrgkanis<\/a>\u00a0(Microsoft Research)<\/p>\n

Adversarial Ranking for Language Generation<\/a>
\nXiaodong He<\/a> (Microsoft Research), Zhengyou Zhang<\/a>\u00a0(Microsoft Research)<\/p>\n

Clustering Billions of Reads for DNA Data Storage<\/a>
\nLuis Ceze,\u00a0Karin Strauss<\/a> (Microsoft Research),\u00a0Sergey Yekhanin<\/a> (Microsoft Research),\u00a0Djordje Jevdjic (Microsoft Research),\u00a0Siena Ang (Microsoft), Konstantin Makarychev (Microsoft)<\/p>\n

Collecting Telemetry Data Privately<\/a>
\nBolin Ding<\/a> (Microsoft Research),\u00a0Janardhan Kulkarni (Microsoft Research),\u00a0Sergey Yekhanin<\/a>\u00a0(Microsoft Research)<\/p>\n

Communication-Efficient Stochastic Gradient Descent, with Applications to Neural Networks<\/a>
\nRyota Tomioka<\/a>\u00a0(Microsoft Research)<\/p>\n

Consistent Robust Regression<\/a>
\nPrateek Jain<\/a>\u00a0(Microsoft Research)<\/p>\n

Decoding with Value Networks for Neural Machine Translation<\/a>
\nDi He<\/a>, Tao Qin<\/a> (Microsoft Research), Tieyan Liu (Microsoft Research)<\/p>\n

Deliberation Networks: Sequence Generation Beyond One-Pass Decoding<\/a>
\nJianxin Lin, Fei Tian<\/a> (Microsoft Research), Tao Qin<\/a> (Microsoft Research), Tie-Yan Liu<\/a>\u00a0(Microsoft Research)<\/p>\n

Efficiency Guarantees from Data<\/a>
\nVasilis Syrgkanis<\/a>\u00a0(Microsoft Research)<\/p>\n

Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach<\/a>
\nHoifung Poon<\/a> (Microsoft Research), Eric Horvitz<\/a>\u00a0(Microsoft Research)<\/p>\n

From Bayesian Sparsity to Gated Recurrent Nets<\/a>
\nDavid Wipf<\/a>\u00a0(Microsoft Research)<\/p>\n

Hybrid Reward Architecture for Reinforcement Learning<\/a>
\nHarm Van Seijen<\/a> (Microsoft Research), Mehdi Fatemi<\/a>\u00a0(Microsoft Research)<\/p>\n

Identifying Outlier Arms in Multi-Armed Bandit<\/a>
\nChi Wang<\/a>\u00a0(Microsoft Research)<\/p>\n

Improving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications<\/a>
\nWei Chen<\/a>\u00a0(Microsoft Research)<\/p>\n

Influence Maximization with \u03b5<\/span><\/span><\/span><\/span>-Almost Submodular Threshold Function<\/a>
\nWei Chen<\/a>\u00a0(Microsoft Research)<\/p>\n

Large-Scale Quadratically Constrained Quadratic Program via Low-Discrepancy Sequences<\/a>
\nKinjal Basu (LinkedIn Corporation),\u00a0Ankan Saha (LinkedIn Corporation), Shaunak Chatterjee (LinkedIn Corporation)<\/p>\n

Learning Mixture of Gaussians with Streaming Data<\/a>
\nPrateek Jain<\/a>\u00a0(Microsoft Research)<\/p>\n

Neural Program Meta-Induction<\/a>
\nJacob Devlin (Microsoft Research), Rishabh Singh<\/a>\u00a0(Microsoft Research), Matthew Hausknecht<\/a>\u00a0(Microsoft Research)<\/p>\n

Off-policy Evaluation for Slate Recommendation<\/a>
\nAdith Swaminathan<\/a> (Microsoft Research),\u00a0Alekh Agarwal<\/a> (Microsoft Research),\u00a0Miro Dudik<\/a> (Microsoft Research),\u00a0Damien Jose (Microsoft),\u00a0Imed Zitouni<\/a>\u00a0(Microsoft Research)<\/p>\n

Online Learning with a Hint<\/a>
\nOfer Dekel<\/a>\u00a0(Microsoft Research)<\/p>\n

Plan, Attend, Generate: Planning for Sequence-to-Sequence Models<\/a>
\nAdam Trischler (Microsoft)<\/p>\n

Q-LDA: Uncovering Latent Patterns in Text-based Sequential Decision Processes<\/a>
\nJianshu Chen<\/a> (Microsoft Research), Lin Xiao<\/a>\u00a0(Microsoft Research)<\/p>\n

Robust Optimization for Non-Convex Objectives<\/a>
\nVasilis Syrgkanis<\/a> (Microsoft Research), Brendan Lucier<\/a>\u00a0(Microsoft Research)<\/p>\n

Stabilizing Training of Generative Adversarial Networks through Regularization<\/a>
\nSebastian Nowozin<\/a> (Microsoft Research)<\/p>\n

Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues<\/a>
\nMoshe Babaioff<\/a>\u00a0(Microsoft Research)<\/p>\n

The Numerics of GANs<\/a>
\nSebastian Nowozin<\/a>\u00a0(Microsoft Research)<\/p>\n

Thy Friend is My Friend: Iterative Collaborative Filtering for Sparse Matrix Estimation<\/a>
\nChristian Borgs<\/a> (Microsoft Research),\u00a0Jennifer Chayes<\/a>\u00a0(Microsoft Research)<\/p>\n

Unsupervised Sequence Classification using Sequential Output Statistics<\/a>
\nJianshu Chen<\/a>\u00a0(Microsoft Research)<\/p>\n

Z-Forcing: Training Stochastic Recurrent Networks<\/a>
\nMarc-Alexandre C\u00f4t\u00e9 (Microsoft Research),\u00a0Alessandro Sordoni\u00a0(Microsoft Research, Maluuba)Opens in a new tab<\/span><\/p>\n

NIPS 2017 Workshops<\/h2>\n

From \u201cWhat if?\u201d To \u201cWhat Next?\u201d: Causal Inference and Machine Learning for Intelligent Decision Making<\/h3>\n

Friday, December 8 @ 9:00 AM\u20136:30 PM | Hall C | Adith Swaminathan<\/a>, Microsoft Research<\/p>\n

This workshop is aimed at facilitating more interactions between researchers in machine learning and causal inference. In particular, it is an opportunity to bring together highly technical individuals who are strongly motivated by the practical importance and real-world impact of their work. Cultivating such interactions will lead to the development of theory, methodology, and – most importantly – practical tools, that better target causal questions across different domains.<\/p>\n

In particular, we will highlight theory, algorithms and applications on automatic decision making systems, such as recommendation engines, medical decision systems and self-driving cars, as both producers and users of data. The challenge here is the feedback between learning from data and then taking actions that may affect what data will be made available for future learning. Learning algorithms have to reason about how changes to the system will affect future data, giving rise to challenging counterfactual and causal reasoning issues that the learning algorithm has to account for. Modern and scalable policy learning algorithms also require operating with non-experimental data, such as logged user interaction data where users click ads suggested by recommender systems trained on historical user clicks.<\/p>\n

To further bring the community together around the use of such interaction data, this workshop will host a Kaggle challenge problem based on the first real-world dataset of logged contextual bandit feedback with non-uniform action-selection propensities. The dataset consists of several gigabytes of data from an ad placement system, which we have processed into multiple well-defined learning problems of increasing complexity, feedback signal, and context. Participants in the challenge problem will be able to discuss their results at the workshop.<\/p>\n

Machine Learning and Computer Security<\/h3>\n

Friday, December 8 @ 9:00 AM\u20135:00 PM | Hyatt Hotel, Shoreline | Donald Brinkman<\/a>, Microsoft Research<\/p>\n

While traditional computer security relies on well-defined attack models and proofs of security, a science of security for machine learning systems has proven more elusive. This is due to a number of obstacles, including (1) the highly varied angles of attack against ML systems, (2) the lack of a clearly defined attack surface (because the source of the data analyzed by ML systems is not easily traced), and (3) the lack of clear formal definitions of security that are appropriate for ML systems. At the same time, security of ML systems is of great import due the recent trend of using ML systems as a line of defense against malicious behavior (e.g., network intrusion, malware, and ransomware), as well as the prevalence of ML systems as parts of sensitive and valuable software systems (e.g., sentiment analyzers for predicting stock prices). This workshop will bring together experts from the computer security and machine learning communities in an attempt to highlight recent work in this area, as well as to clarify the foundations of secure ML and chart out important directions for future work and cross-community collaborations.<\/p>\n

Conversational AI – today’s practice and tomorrow’s potential<\/h3>\n

Friday, December 8 @ 8:00 AM\u20137:00 PM | Grand Ballroom B | Jason Williams<\/a>, Microsoft Research<\/p>\n

This workshop will include invited talks from academia and industry, contributed work, and open discussion. In these talks, senior technical leaders from many of the most popular conversational services will give insights into real usage and challenges at scale. An open call for papers will be issued, and we will prioritize forward-looking papers that propose interesting and impactful contributions. We will end the day with an open discussion, including a panel consisting of academic and industrial researchers.<\/p>\n

Interpreting, Explaining and Visualizing Deep Learning\u2026now what?<\/h3>\n

Saturday, December 9 @ 8:15 AM\u20136:30 PM | Hyatt Regency Ballroom | Hamid Palangi<\/a>, Qiuyuan Huang<\/a>, Paul Smolensky<\/a>, and Xiaodong He<\/a>, Microsoft Research<\/p>\n

Our NIPS 2017 Workshop \u201cInterpreting, Explaining and Visualizing Deep Learning \u2013 Now what?\u201d aims to review recent techniques and establish new theoretical foundations for interpreting and understanding deep learning models. However, it will not stop at the methodological level, but also address the \u201cnow what?\u201d question. This strong focus on the applications of interpretable methods in deep learning distinguishes this workshop from previous events as we aim to take the next step by exploring and extending the practical usefulness of Interpreting, Explaining and Visualizing in Deep Learning. Also with this workshop we aim to identify new fields of applications for interpretable deep learning. Since the workshop will host invited speakers from various application domains (computer vision, NLP, neuroscience, medicine), it will provide an opportunity for participants to learn from each other and initiate new interdisciplinary collaborations. The workshop will contain invited research talks, short methods and applications talks, a poster and demonstration session and a panel discussion. A selection of accepted papers together with the invited contributions will be published in an edited book by Springer LNCS in order to provide a representative overview of recent activities in this emerging research field.<\/p>\n

Co-located workshops<\/h2>\n

Women in Machine Learning<\/h3>\n

Monday, December 4 & Thursday, December 7 @ 2:00 PM\u20132:30 PM | Room 104 | 12th Women in Machine Learning Workshop (WiML 2017)<\/a>, by Hanna Wallach<\/a>, Microsoft Research<\/p>\n

The annual Women in Machine Learning Workshop is the\u00a0flagship event of Women in Machine Learning<\/a>.\u00a0This technical workshop gives female faculty, research scientists, and graduate students in the machine learning community an opportunity to meet, network and exchange ideas, participate in career-focused panel discussions with senior women in industry and academia and learn from each other. Underrepresented minorities and undergraduates interested in machine learning research are encouraged to attend.\u00a0We welcome all genders; however, any formal presentations, i.e. talks and posters, are given\u00a0by women.\u00a0We strive to create an atmosphere in which participants feel comfortable to engage in\u00a0technical and career-related conversations.<\/p>\n

Black in AI<\/h3>\n

Friday, December 8 @ 1:30 PM\u20135:30 PM | Black in AI Workshop @ NIPS 2017<\/a>, by Timnit Gebru<\/a>, Microsoft Research<\/p>\n

The first Black in AI event will be co-located with NIPS 2017. The goal is to gather people in the field to share ideas and discuss\u00a0initiatives to increase the presence of Black\u00a0people in the field of\u00a0artificial intelligence, for both diversity and data bias prevention purposes. At this workshop, Black researchers in AI will also have the opportunity to present their work during our oral and poster sessions.Opens in a new tab<\/span><\/p>\n","tab-content":[{"id":0,"name":"About","content":"Microsoft is excited to be a Platinum sponsor of the thirty-first annual conference on Neural Information Processing Systems (NIPS)<\/a>. Over 80 of our researchers are involved in spotlight sessions, presentations, symposiums, posters, accepted papers, and workshops at NIPS (see schedule below). Stop by our booth (#315, Exhibit Hall B) to see HoloLens<\/a> and Windows Mixed Reality<\/a> in action, as well as to find out about career opportunities at Microsoft and enter for your chance to win an Xbox One X<\/a> gaming console. Follow @MSFTResearch<\/a> for the latest information coming from the event.\r\n

NIPS organizing committee<\/h2>\r\nHanna Wallach<\/a>, Program Co-chair\r\nJenn Wortman Vaughan<\/a>, Tutorial chair\r\nMarkus Weimer<\/a>,\u00a0Demonstration and competition chair\r\n

Workshop organizers<\/h2>\r\nEvelyne Viegas<\/a>, Machine Learning Challenges as a Research Tool<\/a>\r\nNicolo Fusi<\/a>, Machine Learning in Computational Biology<\/a>\r\nSiddhartha Sen<\/a>, ML Systems Workshop<\/a>\r\nAlekh Agarwal<\/a>, OPT 2017: Optimization for Machine Learning<\/a>\r\nJennifer Wortman Vaughan<\/a>, Learning in the Presence of Strategic Behavior<\/a>\r\nManik Varma<\/a>, Extreme Classification: Multi-class & Multi-label Learning in Extremely Large Label Spaces<\/a>\r\nVasilis Syrgkanis<\/a>, Learning in the Presence of Strategic Behavior<\/a>\r\n

Symposium\u00a0organizers<\/h2>\r\nPatrice Simard<\/a>, Interpretable Machine Learning<\/a>\r\nRich Caruana<\/a>, Interpretable Machine Learning<\/a>\r\n

Invited speaker<\/h2>\r\nDecember 5 @ 1:50\u20132:40 PM | Kate Crawford<\/span><\/a>, <\/span>The Trouble with Bias<\/span><\/a>\r\n

Careers at Microsoft information session<\/h2>\r\nDecember 6 @ 12:30\u20131:20 PM |\u00a0Eric Horvitz<\/a>, Christopher Bishop<\/a>, Jennifer Chayes<\/a>, and Mir Rosenberg\r\n

Spotlight sessions<\/h2>\r\nDecember 5 @ 3:30\u20133:35 PM | Clustering Billions of Reads for DNA Data Storage<\/a>\r\nCyrus Rashtchian, Konstantin Makarychev, Luis Ceze, Karin Strauss<\/a>, Sergey Yekhanin<\/a>, Djordje Jevdjic, Miklos Racz, and Siena Ang\r\n\r\nDecember 6 @ 11:20\u201311:25 AM | Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues\r\n<\/a>Noga Alon, Moshe Babaioff<\/a>, Yannai A. Gonczarowski, Yishay Mansour, Shay Moran, and Amir Yehudayoff\r\n\r\nDecember 6 @ 5:15\u20135:20 PM | Repeated Inverse Reinforcement Learning<\/a>\r\nKareem Amin, Nan Jiang<\/a>, and Satinder Singh\r\n

Oral presentations<\/h2>\r\nDecember 5 @ 10:55\u201311:00 AM | Robust Optimization for Non-Convex Objectives\r\n<\/a>Yaron Singer, Robert S Chen, Vasilis Syrgkanis<\/a>, and\u00a0Brendan Lucier<\/a>\r\n\r\nDecember 6 @ 4:20\u20134:35 PM | Off-policy evaluation for slate recommendation<\/a>\r\nAdith Swaminathan<\/a>, Akshay Krishnamurthy, Alekh Agarwal<\/a>, Miro Dudik<\/a>, John Langford<\/a>, Damien Jose, and\u00a0Imed Zitouni<\/a>"},{"id":1,"name":"Accepted Papers","content":"\"A Decomposition of Forecast Error in Prediction Markets<\/a>\" by\u00a0Miro Dudik<\/a>, Sebastien Lahaie, Ryan M Rogers, and\u00a0Jennifer Wortman Vaughan<\/a>\r\n\r\n\"A Highly Efficient Gradient Boosting Decision Tree<\/a>\" by\u00a0Guolin Ke, Qi Meng, Taifeng Wang<\/a>, Wei Chen<\/a>, Weidong Ma<\/a>, and\u00a0Tie-Yan Liu<\/a>\r\n\r\n\"A Sample Complexity Measure with Applications to Learning Optimal Auctions<\/a>\" by\u00a0Vasilis Syrgkanis<\/a>\r\n\r\n\"Accuracy First: Selecting a Differential Privacy Level for Accuracy Constrained ERM<\/a>\" by\u00a0Steven Wu<\/a>, Bo Waggoner, Seth Neel, Aaron Roth, and Katrina Ligett\r\n\r\n\"Adversarial Ranking for Language Generation<\/a>\" by\u00a0Dianqi Li, Kevin Lin, Xiaodong He<\/a>, Ming-ting Sun, and\u00a0Zhengyou Zhang<\/a>\r\n\r\n\"Clustering Billions of Reads for DNA Data Storage<\/a>\" by\u00a0Cyrus Rashtchian, Konstantin Makarychev, Luis Ceze, Karin Strauss<\/a>, Sergey Yekhanin<\/a>, Djordje Jevdjic, Miklos Racz, and Siena Ang\r\n\r\n\"Collecting Telemetry Data Privately<\/a>\" by\u00a0Bolin Ding<\/a>, Janardhan Kulkarni, and\u00a0Sergey Yekhanin<\/a>\r\n\r\n\"Consistent Robust Regression<\/a>\" by\u00a0Kush Bhatia, Prateek Jain<\/a>, and Purushottam Kar\r\n\r\n\"Decoding with Value Networks for Neural Machine Translation<\/a>\" by\u00a0Di He<\/a>, Hanqing Lu, Yingce Xia, Tao Qin<\/a>, Liwei Wang, and Tieyan Liu\r\n\r\n\"Deliberation Networks: Sequence Generation Beyond One-Pass Decoding<\/a>\" by\u00a0Yingce Xia, Lijun Wu, Jianxin Lin, Fei Tian<\/a>, Tao Qin<\/a>, and\u00a0Tie-Yan Liu<\/a>\r\n\r\n\"Efficiency Guarantees from Data<\/a>\" by\u00a0Darrell Hoy, Tremor Technologies; Denis Nekipelov, University of Virginia; and\u00a0Vasilis Syrgkanis<\/a>, Microsoft Research\r\n\r\n\"Estimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach<\/a>\" by\u00a0Emmanouil Platanios, Carnegie Mellon University;\u00a0Hoifung Poon<\/a>, Microsoft Research; Tom M. Mitchell, Carnegie Mellon University; and\u00a0Eric J. Horvitz<\/a>, Microsoft Research\r\n\r\n\"From Bayesian Sparsity to Gated Recurrent Nets<\/a>\" by\u00a0Hao He, Massachusetts Institute of Technology;\u00a0Bo Xin<\/a>, Microsoft Research; and\u00a0David Wipf<\/a>, Microsoft Research\r\n\r\n\"Hybrid Reward Architecture for Reinforcement Learning<\/a>\" by\u00a0Harm Van Seijen<\/a>, Microsoft Research;\u00a0Romain Laroche<\/a>, Microsoft Research, Maluuba;\u00a0Mehdi Fatemi<\/a>, Microsoft Research; and Joshua Romoff, McGill University\r\n\r\n\"Identifying Outlier Arms in Multi-Armed Bandit<\/a>\" by\u00a0Honglei Zhuang, University of Illinois;\u00a0Chi Wang<\/a>, Microsoft Research; and Yifan Wang, Tsinghua University\r\n\r\n\"Improving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications<\/a>\" by\u00a0Qinshi Wang and\u00a0Wei Chen<\/a>\r\n\r\n\"Inference in Graphical Models via Semidefinite Programming Hierarchies<\/a>\" by\u00a0Murat Erdogdu<\/a>, Yash Deshpande, and Andrea Montanari\r\n\r\n\"Influence Maximization with \u03b5<\/span><\/span><\/span><\/span>-Almost Submodular Threshold Function<\/a>\" by Qiang Li, Institute of Computing Technol;\u00a0Wei Chen<\/a>, Microsoft Research; Xiaoming Sun, Institute of Computing Technology, Chinese Academy of Sciences; and Jialin Zhang, Institute of Computing Technology, Chinese Academy of Sciences\r\n\r\n\"Large-Scale Quadratically Constrained Quadratic Program via Low-Discrepancy Sequences<\/a>\" by\u00a0Kinjal Basu,\u00a0Ankan Saha, and\u00a0Shaunak Chatterjee,\u00a0LinkedIn Corporation\r\n\r\n\"Learning Mixture of Gaussians with Streaming Data<\/a>\" by\u00a0Aditi Raghunathan, Stanford University;\u00a0Prateek Jain<\/a>, Microsoft Research; and Ravishankar Krishnawamy, Microsoft Research\r\n\r\n\"Linear Convergence of a Frank-Wolfe Type Algorithm over Trace-Norm Balls<\/a>\" by\u00a0Zeyuan Allen-Zhu<\/a>, Microsoft Research; Elad Hazan, Princeton University; Wei Hu, Princeton University; Yuanzhi Li, Princeton University\r\n\r\n\"The Importance of Communities for Learning to Influence<\/a>\" by\u00a0Eric Balkanski, Harvard University;\u00a0Nicole Immorlica<\/a>, Microsoft Research; and Yaron Singer, Harvard University\r\n\r\n\"Mean Field Residual Networks: On the Edge of Chaos<\/a>\" by Greg Yang, Microsoft Research; Samuel S. Schoenholz, Google Brain\r\n\r\n\"Multi-Task Learning for Contextual Bandits<\/a>\" by\u00a0Aniket Anand Deshmukh, University of Michigan, Ann Arbor; Urun Dogan, Microsoft; and Clay Scott, University of Michigan\r\n\r\n\"Neural Program Meta-Induction<\/a>\" by\u00a0Jacob Devlin, Microsoft Research; Rudy R Bunel, Oxford University;\u00a0Rishabh Singh<\/a>, Microsoft Research;\u00a0Matthew Hausknecht<\/a>, Microsoft Research; and Pushmeet Kohli, DeepMind\r\n\r\n\"Non-convex Robust PCA<\/a>\" by Praneeth Netrapalli, Microsoft Research; Niranjan Uma Naresh, UC Irvine; Sujay Sanghavi, UT-Austin; Animashree Anadkumar, UC-Irvine; Prateek Jain, Microsoft Research\r\n\r\n\"Off-policy Evaluation for Slate Recommendation<\/a>\" by\u00a0Adith Swaminathan<\/a>, Microsoft Research; Akshay Krishnamurthy, University of Massachusetts;\u00a0Alekh Agarwal<\/a>, Microsoft Research;\u00a0Miro Dudik<\/a>, Microsoft Research;\u00a0John Langford<\/a>, Microsoft Research; Damien Jose, Microsoft; and\u00a0Imed Zitouni<\/a>, Microsoft Research\r\n\r\n\"Online Learning with a Hint<\/a>\" by\u00a0Ofer Dekel<\/a>, Microsoft Research; Arthur Flajolet, Massachusetts Institute of Technology; Nika Haghtalab, Carnegie Mellon University; and Patrick Jaillet, Massachusetts Institute of Technology\r\n\r\n\"Plan, Attend, Generate: Planning for Sequence-to-Sequence Models<\/a>\" by\u00a0Caglar Gulcehre, Deepmind; Francis Dutil, MILA; Adam Trischler, Microsoft; and Yoshua Bengio, University of Montreal\r\n\r\n\"Q-LDA: Uncovering Latent Patterns in Text-based Sequential Decision Processes<\/a>\" by\u00a0Jianshu Chen<\/a>, Microsoft Research; Chong Wang, Princeton University;\u00a0Lin Xiao<\/a>, Microsoft Research; Ji He, University Washington;\u00a0Lihong Li<\/a>, Microsoft Research; and Li Deng, Citadel\r\n\r\n\"QSGD: Communication-Efficient Stochastic Gradient Descent, with Applications to Neural Networks<\/a>\" by\u00a0Dan Alistarh, Demjan Grubic, Jerry Li, Ryota Tomioka<\/a>, and Milan Vojnovic\r\n\r\n\"Repeated Inverse Reinforcement Learning<\/a>\" by\u00a0Kareem Amin, Google Research;\u00a0Nan Jiang<\/a>, Microsoft Research; and Satinder Singh, University of Michigan\r\n\r\n\"Robust Estimation of Neural Signals in Calcium Imaging<\/a>\" by\u00a0Hakan Inan, Stanford University;\u00a0Murat Erdogdu<\/a>, Microsoft Research; and Mark Schnitzer, Stanford University\r\n\r\n\"Robust Optimization for Non-Convex Objectives<\/a>\" by\u00a0Yaron Singer, Harvard University; Robert S Chen, Harvard University;\u00a0Vasilis Syrgkanis<\/a>, Microsoft Research; and\u00a0Brendan Lucier<\/a>, Microsoft Research\r\n\r\n\"Stabilizing Training of Generative Adversarial Networks through Regularization<\/a>\" by\u00a0Kevin Roth, ETH; Aurelien Lucchi, ETH Zurich;\u00a0Sebastian Nowozin<\/a>, Microsoft Research; and Thomas Hofmann, ETH Zurich\r\n\r\n\"Submultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues<\/a>\" by\u00a0Noga Alon, Tel Aviv University;\u00a0Moshe Babaioff<\/a>, Microsoft Research; Yannai A. Gonczarowski, The Hebrew University of Jerusalem and Microsoft Research; Yishay Mansour, Tel Aviv University; Shay Moran, IAS, Princeton; and Amir Yehudayoff, Technion - Israel Institute of Technology\r\n\r\n\"The Numerics of GANs<\/a>\" by\u00a0Lars Mescheder, Max-Planck Institute Tuebingen;\u00a0Sebastian Nowozin<\/a>, Microsoft Research; and Andreas Geiger, MPI T\u00fcbingen\r\n\r\n\"Thy Friend is My Friend: Iterative Collaborative Filtering for Sparse Matrix Estimation<\/a>\" by\u00a0Christian Borgs<\/a>, Microsoft Research;\u00a0Jennifer Chayes<\/a>, Microsoft Research;\u00a0Christina Lee<\/a>, Microsoft Research; and Devavrat Shah, Massachusetts Institute of Technology\r\n\r\n\"Unsupervised Sequence Classification using Sequential Output Statistics<\/a>\" by\u00a0Yu Liu, SUNY Buffalo;\u00a0Jianshu Chen<\/a>, Microsoft Research; and Li Deng, Citadel\r\n\r\n\"Z-Forcing: Training Stochastic Recurrent Networks<\/a>\" by\u00a0Marc-Alexandre C\u00f4t\u00e9, Microsoft Research; Alessandro Sordoni, Microsoft Research, Maluuba; Anirudh Goyal, Universit\u00e9 de Montr\u00e9al; Nan Ke, MILA, \u00c9cole Polytechnique de Montr\u00e9al; and Yoshua Bengio, University of Montreal"},{"id":2,"name":"Posters","content":"A Decomposition of Forecast Error in Prediction Markets<\/a>\r\nMiro Dudik<\/a> (Microsoft Research),\u00a0Jennifer Wortman Vaughan<\/a>\u00a0(Microsoft Research)\r\n\r\nA Highly Efficient Gradient Boosting Decision Tree<\/a>\r\nGuolin Ke (Microsoft Research), Taifeng Wang<\/a> (Microsoft Research), Wei Chen<\/a> (Microsoft Research), Weidong Ma<\/a> (Microsoft Research), Tie-Yan Liu<\/a>\u00a0(Microsoft Research)\r\n\r\nA Sample Complexity Measure with Applications to Learning Optimal Auctions<\/a>\r\nVasilis Syrgkanis<\/a>\u00a0(Microsoft Research)\r\n\r\nAdversarial Ranking for Language Generation<\/a>\r\nXiaodong He<\/a> (Microsoft Research), Zhengyou Zhang<\/a>\u00a0(Microsoft Research)\r\n\r\nClustering Billions of Reads for DNA Data Storage<\/a>\r\nLuis Ceze,\u00a0Karin Strauss<\/a> (Microsoft Research),\u00a0Sergey Yekhanin<\/a> (Microsoft Research),\u00a0Djordje Jevdjic (Microsoft Research),\u00a0Siena Ang (Microsoft), Konstantin Makarychev (Microsoft)\r\n\r\nCollecting Telemetry Data Privately<\/a>\r\nBolin Ding<\/a> (Microsoft Research),\u00a0Janardhan Kulkarni (Microsoft Research),\u00a0Sergey Yekhanin<\/a>\u00a0(Microsoft Research)\r\n\r\nCommunication-Efficient Stochastic Gradient Descent, with Applications to Neural Networks<\/a>\r\nRyota Tomioka<\/a>\u00a0(Microsoft Research)\r\n\r\nConsistent Robust Regression<\/a>\r\nPrateek Jain<\/a>\u00a0(Microsoft Research)\r\n\r\nDecoding with Value Networks for Neural Machine Translation<\/a>\r\nDi He<\/a>, Tao Qin<\/a> (Microsoft Research), Tieyan Liu (Microsoft Research)\r\n\r\nDeliberation Networks: Sequence Generation Beyond One-Pass Decoding<\/a>\r\nJianxin Lin, Fei Tian<\/a> (Microsoft Research), Tao Qin<\/a> (Microsoft Research), Tie-Yan Liu<\/a>\u00a0(Microsoft Research)\r\n\r\nEfficiency Guarantees from Data<\/a>\r\nVasilis Syrgkanis<\/a>\u00a0(Microsoft Research)\r\n\r\nEstimating Accuracy from Unlabeled Data: A Probabilistic Logic Approach<\/a>\r\nHoifung Poon<\/a> (Microsoft Research), Eric Horvitz<\/a>\u00a0(Microsoft Research)\r\n\r\nFrom Bayesian Sparsity to Gated Recurrent Nets<\/a>\r\nDavid Wipf<\/a>\u00a0(Microsoft Research)\r\n\r\nHybrid Reward Architecture for Reinforcement Learning<\/a>\r\nHarm Van Seijen<\/a> (Microsoft Research), Mehdi Fatemi<\/a>\u00a0(Microsoft Research)\r\n\r\nIdentifying Outlier Arms in Multi-Armed Bandit<\/a>\r\nChi Wang<\/a>\u00a0(Microsoft Research)\r\n\r\nImproving Regret Bounds for Combinatorial Semi-Bandits with Probabilistically Triggered Arms and Its Applications<\/a>\r\nWei Chen<\/a>\u00a0(Microsoft Research)\r\n\r\nInfluence Maximization with \u03b5<\/span><\/span><\/span><\/span>-Almost Submodular Threshold Function<\/a>\r\nWei Chen<\/a>\u00a0(Microsoft Research)\r\n\r\nLarge-Scale Quadratically Constrained Quadratic Program via Low-Discrepancy Sequences<\/a>\r\nKinjal Basu (LinkedIn Corporation),\u00a0Ankan Saha (LinkedIn Corporation), Shaunak Chatterjee (LinkedIn Corporation)\r\n\r\nLearning Mixture of Gaussians with Streaming Data<\/a>\r\nPrateek Jain<\/a>\u00a0(Microsoft Research)\r\n\r\nNeural Program Meta-Induction<\/a>\r\nJacob Devlin (Microsoft Research), Rishabh Singh<\/a>\u00a0(Microsoft Research), Matthew Hausknecht<\/a>\u00a0(Microsoft Research)\r\n\r\nOff-policy Evaluation for Slate Recommendation<\/a>\r\nAdith Swaminathan<\/a> (Microsoft Research),\u00a0Alekh Agarwal<\/a> (Microsoft Research),\u00a0Miro Dudik<\/a> (Microsoft Research),\u00a0Damien Jose (Microsoft),\u00a0Imed Zitouni<\/a>\u00a0(Microsoft Research)\r\n\r\nOnline Learning with a Hint<\/a>\r\nOfer Dekel<\/a>\u00a0(Microsoft Research)\r\n\r\nPlan, Attend, Generate: Planning for Sequence-to-Sequence Models<\/a>\r\nAdam Trischler (Microsoft)\r\n\r\nQ-LDA: Uncovering Latent Patterns in Text-based Sequential Decision Processes<\/a>\r\nJianshu Chen<\/a> (Microsoft Research), Lin Xiao<\/a>\u00a0(Microsoft Research)\r\n\r\nRobust Optimization for Non-Convex Objectives<\/a>\r\nVasilis Syrgkanis<\/a> (Microsoft Research), Brendan Lucier<\/a>\u00a0(Microsoft Research)\r\n\r\nStabilizing Training of Generative Adversarial Networks through Regularization<\/a>\r\nSebastian Nowozin<\/a> (Microsoft Research)\r\n\r\nSubmultiplicative Glivenko-Cantelli and Uniform Convergence of Revenues<\/a>\r\nMoshe Babaioff<\/a>\u00a0(Microsoft Research)\r\n\r\nThe Numerics of GANs<\/a>\r\nSebastian Nowozin<\/a>\u00a0(Microsoft Research)\r\n\r\nThy Friend is My Friend: Iterative Collaborative Filtering for Sparse Matrix Estimation<\/a>\r\nChristian Borgs<\/a> (Microsoft Research),\u00a0Jennifer Chayes<\/a>\u00a0(Microsoft Research)\r\n\r\nUnsupervised Sequence Classification using Sequential Output Statistics<\/a>\r\nJianshu Chen<\/a>\u00a0(Microsoft Research)\r\n\r\nZ-Forcing: Training Stochastic Recurrent Networks<\/a>\r\nMarc-Alexandre C\u00f4t\u00e9 (Microsoft Research),\u00a0Alessandro Sordoni\u00a0(Microsoft Research, Maluuba)"},{"id":3,"name":"Workshops","content":"

NIPS 2017 Workshops<\/h2>\r\n

From \u201cWhat if?\u201d To \u201cWhat Next?\u201d: Causal Inference and Machine Learning for Intelligent Decision Making<\/h3>\r\nFriday, December 8 @ 9:00 AM\u20136:30 PM | Hall C | Adith Swaminathan<\/a>, Microsoft Research\r\n\r\nThis workshop is aimed at facilitating more interactions between researchers in machine learning and causal inference. In particular, it is an opportunity to bring together highly technical individuals who are strongly motivated by the practical importance and real-world impact of their work. Cultivating such interactions will lead to the development of theory, methodology, and - most importantly - practical tools, that better target causal questions across different domains.\r\n\r\nIn particular, we will highlight theory, algorithms and applications on automatic decision making systems, such as recommendation engines, medical decision systems and self-driving cars, as both producers and users of data. The challenge here is the feedback between learning from data and then taking actions that may affect what data will be made available for future learning. Learning algorithms have to reason about how changes to the system will affect future data, giving rise to challenging counterfactual and causal reasoning issues that the learning algorithm has to account for. Modern and scalable policy learning algorithms also require operating with non-experimental data, such as logged user interaction data where users click ads suggested by recommender systems trained on historical user clicks.\r\n\r\nTo further bring the community together around the use of such interaction data, this workshop will host a Kaggle challenge problem based on the first real-world dataset of logged contextual bandit feedback with non-uniform action-selection propensities. The dataset consists of several gigabytes of data from an ad placement system, which we have processed into multiple well-defined learning problems of increasing complexity, feedback signal, and context. Participants in the challenge problem will be able to discuss their results at the workshop.\r\n

Machine Learning and Computer Security<\/h3>\r\nFriday, December 8 @ 9:00 AM\u20135:00 PM | Hyatt Hotel, Shoreline | Donald Brinkman<\/a>, Microsoft Research\r\n\r\nWhile traditional computer security relies on well-defined attack models and proofs of security, a science of security for machine learning systems has proven more elusive. This is due to a number of obstacles, including (1) the highly varied angles of attack against ML systems, (2) the lack of a clearly defined attack surface (because the source of the data analyzed by ML systems is not easily traced), and (3) the lack of clear formal definitions of security that are appropriate for ML systems. At the same time, security of ML systems is of great import due the recent trend of using ML systems as a line of defense against malicious behavior (e.g., network intrusion, malware, and ransomware), as well as the prevalence of ML systems as parts of sensitive and valuable software systems (e.g., sentiment analyzers for predicting stock prices). This workshop will bring together experts from the computer security and machine learning communities in an attempt to highlight recent work in this area, as well as to clarify the foundations of secure ML and chart out important directions for future work and cross-community collaborations.\r\n

Conversational AI - today's practice and tomorrow's potential<\/h3>\r\nFriday, December 8 @ 8:00 AM\u20137:00 PM | Grand Ballroom B | Jason Williams<\/a>, Microsoft Research\r\n\r\nThis workshop will include invited talks from academia and industry, contributed work, and open discussion. In these talks, senior technical leaders from many of the most popular conversational services will give insights into real usage and challenges at scale. An open call for papers will be issued, and we will prioritize forward-looking papers that propose interesting and impactful contributions. We will end the day with an open discussion, including a panel consisting of academic and industrial researchers.\r\n

Interpreting, Explaining and Visualizing Deep Learning\u2026now what?<\/h3>\r\nSaturday, December 9 @ 8:15 AM\u20136:30 PM | Hyatt Regency Ballroom | Hamid Palangi<\/a>, Qiuyuan Huang<\/a>, Paul Smolensky<\/a>, and Xiaodong He<\/a>, Microsoft Research\r\n\r\nOur NIPS 2017 Workshop \u201cInterpreting, Explaining and Visualizing Deep Learning \u2013 Now what?\u201d aims to review recent techniques and establish new theoretical foundations for interpreting and understanding deep learning models. However, it will not stop at the methodological level, but also address the \u201cnow what?\u201d question. This strong focus on the applications of interpretable methods in deep learning distinguishes this workshop from previous events as we aim to take the next step by exploring and extending the practical usefulness of Interpreting, Explaining and Visualizing in Deep Learning. Also with this workshop we aim to identify new fields of applications for interpretable deep learning. Since the workshop will host invited speakers from various application domains (computer vision, NLP, neuroscience, medicine), it will provide an opportunity for participants to learn from each other and initiate new interdisciplinary collaborations. The workshop will contain invited research talks, short methods and applications talks, a poster and demonstration session and a panel discussion. A selection of accepted papers together with the invited contributions will be published in an edited book by Springer LNCS in order to provide a representative overview of recent activities in this emerging research field.\r\n

Co-located workshops<\/h2>\r\n

Women in Machine Learning<\/h3>\r\nMonday, December 4 & Thursday, December 7 @ 2:00 PM\u20132:30 PM | Room 104 | 12th Women in Machine Learning Workshop (WiML 2017)<\/a>, by Hanna Wallach<\/a>, Microsoft Research\r\n\r\nThe annual Women in Machine Learning Workshop is the\u00a0flagship event of Women in Machine Learning<\/a>.\u00a0This technical workshop gives female faculty, research scientists, and graduate students in the machine learning community an opportunity to meet, network and exchange ideas, participate in career-focused panel discussions with senior women in industry and academia and learn from each other. Underrepresented minorities and undergraduates interested in machine learning research are encouraged to attend.\u00a0We welcome all genders; however, any formal presentations, i.e. talks and posters, are given\u00a0by women.\u00a0We strive to create an atmosphere in which participants feel comfortable to engage in\u00a0technical and career-related conversations.\r\n

Black in AI<\/h3>\r\nFriday, December 8 @ 1:30 PM\u20135:30 PM | Black in AI Workshop @ NIPS 2017<\/a>, by Timnit Gebru<\/a>, Microsoft Research\r\n\r\nThe first Black in AI event will be co-located with NIPS 2017. The goal is to gather people in the field to share ideas and discuss\u00a0initiatives to increase the presence of Black\u00a0people in the field of\u00a0artificial intelligence, for both diversity and data bias prevention purposes. At this workshop, Black researchers in AI will also have the opportunity to present their work during our oral and poster sessions."}],"msr_startdate":"2017-12-04","msr_enddate":"2017-12-09","msr_event_time":"","msr_location":"Long Beach, California","msr_event_link":"","msr_event_recording_link":"","msr_startdate_formatted":"December 4, 2017","msr_register_text":"Watch now","msr_cta_link":"","msr_cta_text":"","msr_cta_bi_name":"","featured_image_thumbnail":"\"NIPS","event_excerpt":"The Thirty-first Annual Conference on Neural Information Processing Systems (NIPS) is a multi-track machine learning and computational neuroscience conference that includes invited talks, demonstrations, symposia and oral and poster presentations of refereed papers.","msr_research_lab":[],"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\/425610","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":17,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/425610\/revisions"}],"predecessor-version":[{"id":1147143,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event\/425610\/revisions\/1147143"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/446514"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=425610"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=425610"},{"taxonomy":"msr-region","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-region?post=425610"},{"taxonomy":"msr-event-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-event-type?post=425610"},{"taxonomy":"msr-video-type","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-video-type?post=425610"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=425610"},{"taxonomy":"msr-program-audience","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-program-audience?post=425610"},{"taxonomy":"msr-post-option","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-post-option?post=425610"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=425610"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}