\n \n \n
\n
\n\n \n \n \n \n \n \n Causal Effect Inference with Deep Latent-Variable Models.\n \n \n \n \n\n\n \n Louizos, C.; Shalit, U.; Mooij, J.; Sontag, D.; Zemel, R. S.; and Welling, M.\n\n\n \n\n\n\n In
Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS), of
NIPS'17, 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 14 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@inproceedings{LouizosEtAl_nips17,\n author = {Christos Louizos and\n Uri Shalit and\n Joris Mooij and\n David Sontag and\n Richard S. Zemel and\n Max Welling},\n title = {Causal Effect Inference with Deep Latent-Variable Models},\n booktitle = {Proceedings of the 31st International Conference on Neural Information Processing Systems (NIPS)},\n series = {NIPS'17},\n year = {2017},\n keywords = {Machine learning, Causal inference, Deep learning},\n url_Paper = {https://arxiv.org/pdf/1705.08821.pdf},\n abstract = {Learning individual-level causal effects from observational data, such as inferring the most effective medication for a specific patient, is a problem of growing importance for policy makers. The most important aspect of inferring causal effects from observational data is the handling of confounders, factors that affect both an intervention and its outcome. A carefully designed observational study attempts to measure all important confounders. However, even if one does not have direct access to all confounders, there may exist noisy and uncertain measurement of proxies for confounders. We build on recent advances in latent variable modelling to simultaneously estimate the unknown latent space summarizing the confounders and the causal effect. Our method is based on Variational Autoencoders (VAE) which follow the causal structure of inference with proxies. We show our method is significantly more robust than existing methods, and matches the state-of-the-art on previous benchmarks focused on individual treatment effects.}\n}\n\n
\n
\n\n\n
\n Learning individual-level causal effects from observational data, such as inferring the most effective medication for a specific patient, is a problem of growing importance for policy makers. The most important aspect of inferring causal effects from observational data is the handling of confounders, factors that affect both an intervention and its outcome. A carefully designed observational study attempts to measure all important confounders. However, even if one does not have direct access to all confounders, there may exist noisy and uncertain measurement of proxies for confounders. We build on recent advances in latent variable modelling to simultaneously estimate the unknown latent space summarizing the confounders and the causal effect. Our method is based on Variational Autoencoders (VAE) which follow the causal structure of inference with proxies. We show our method is significantly more robust than existing methods, and matches the state-of-the-art on previous benchmarks focused on individual treatment effects.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Estimating individual treatment effect: generalization bounds and algorithms.\n \n \n \n \n\n\n \n Shalit, U.; Johansson, F. D.; and Sontag, D.\n\n\n \n\n\n\n In
Proceedings of the 34th International Conference on Machine Learning (ICML), pages 3076–3085, 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 9 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@inproceedings{ShalitEtAl_icml17,\n author = {Uri Shalit and\n Fredrik D. Johansson and\n David Sontag},\n title = {Estimating individual treatment effect: generalization bounds and\n algorithms},\n booktitle = {Proceedings of the 34th International Conference on Machine Learning (ICML)},\n pages = {3076--3085},\n year = {2017},\n keywords = {Machine learning, Causal inference, Deep learning},\n url_Paper = {http://arxiv.org/pdf/1606.03976.pdf},\n abstract = {There is intense interest in applying machine learning to problems of causal inference in fields such as healthcare, economics and education. In particular, individual-level causal inference has important applications such as precision medicine. We give a new theoretical analysis and family of algorithms for predicting individual treatment effect (ITE) from observational data, under the assumption known as strong ignorability. The algorithms learn a "balanced" representation such that the induced treated and control distributions look similar. We give a novel, simple and intuitive generalization-error bound showing that the expected ITE estimation error of a representation is bounded by a sum of the standard generalization-error of that representation and the distance between the treated and control distributions induced by the representation. We use Integral Probability Metrics to measure distances between distributions, deriving explicit bounds for the Wasserstein and Maximum Mean Discrepancy (MMD) distances. Experiments on real and simulated data show the new algorithms match or outperform the state-of-the-art.}\n}\n\n
\n
\n\n\n
\n There is intense interest in applying machine learning to problems of causal inference in fields such as healthcare, economics and education. In particular, individual-level causal inference has important applications such as precision medicine. We give a new theoretical analysis and family of algorithms for predicting individual treatment effect (ITE) from observational data, under the assumption known as strong ignorability. The algorithms learn a \"balanced\" representation such that the induced treated and control distributions look similar. We give a novel, simple and intuitive generalization-error bound showing that the expected ITE estimation error of a representation is bounded by a sum of the standard generalization-error of that representation and the distance between the treated and control distributions induced by the representation. We use Integral Probability Metrics to measure distances between distributions, deriving explicit bounds for the Wasserstein and Maximum Mean Discrepancy (MMD) distances. Experiments on real and simulated data show the new algorithms match or outperform the state-of-the-art.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Simultaneous Learning of Trees and Representations for Extreme Classification and Density Estimation.\n \n \n \n \n\n\n \n Jernite, Y.; Choromanska, A.; and Sontag, D.\n\n\n \n\n\n\n In
Proceedings of the 34th International Conference on Machine Learning (ICML), pages 1665–1674, 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@inproceedings{JerniteEtAl_icml17,\n author = {Yacine Jernite and\n Anna Choromanska and\n David Sontag},\n title = {Simultaneous Learning of Trees and Representations for Extreme Classification\n and Density Estimation},\n booktitle = {Proceedings of the 34th International Conference on Machine Learning (ICML)},\n pages = {1665--1674},\n year = {2017},\n keywords = {Machine learning, Natural language processing, Deep learning},\n url_Paper = {https://arxiv.org/pdf/1610.04658.pdf},\n abstract = {We consider multi-class classification where the predictor has a hierarchical structure that allows for a very large number of labels both at train and test time. The predictive power of such models can heavily depend on the structure of the tree, and although past work showed how to learn the tree structure, it expected that the feature vectors remained static. We provide a novel algorithm to simultaneously perform representation learning for the input data and learning of the hierarchical predictor. Our approach optimizes an objective function which favors balanced and easily-separable multi-way node partitions. We theoretically analyze this objective, showing that it gives rise to a boosting style property and a bound on classification error. We next show how to extend the algorithm to conditional density estimation. We empirically validate both variants of the algorithm on text classification and language modeling, respectively, and show that they compare favorably to common baselines in terms of accuracy and running time.}\n}\n\n
\n
\n\n\n
\n We consider multi-class classification where the predictor has a hierarchical structure that allows for a very large number of labels both at train and test time. The predictive power of such models can heavily depend on the structure of the tree, and although past work showed how to learn the tree structure, it expected that the feature vectors remained static. We provide a novel algorithm to simultaneously perform representation learning for the input data and learning of the hierarchical predictor. Our approach optimizes an objective function which favors balanced and easily-separable multi-way node partitions. We theoretically analyze this objective, showing that it gives rise to a boosting style property and a bound on classification error. We next show how to extend the algorithm to conditional density estimation. We empirically validate both variants of the algorithm on text classification and language modeling, respectively, and show that they compare favorably to common baselines in terms of accuracy and running time.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Structured Inference Networks for Nonlinear State Space Models.\n \n \n \n \n\n\n \n Krishnan, R. G.; Shalit, U.; and Sontag, D.\n\n\n \n\n\n\n In
Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence, pages 2101–2109, 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n \n \n 3 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@inproceedings{KrishnanEtAl_aaai17,\n author = {Rahul G. Krishnan and\n Uri Shalit and\n David Sontag},\n title = {Structured Inference Networks for Nonlinear State Space Models},\n booktitle = {Proceedings of the Thirty-First {AAAI} Conference on Artificial Intelligence},\n pages = {2101--2109},\n year = {2017},\n keywords = {Machine learning, Unsupervised learning, Deep learning, Health care, Approximate inference in graphical models},\n url_Paper = {https://arxiv.org/pdf/1609.09869.pdf},\n abstract = {Gaussian state space models have been used for decades as generative models of sequential data. They admit an intuitive probabilistic interpretation, have a simple functional form, and enjoy widespread adoption. We introduce a unified algorithm to efficiently learn a broad class of linear and non-linear state space models, including variants where the emission and transition distributions are modeled by deep neural networks. Our learning algorithm simultaneously learns a compiled inference network and the generative model, leveraging a structured variational approximation parameterized by recurrent neural networks to mimic the posterior distribution. We apply the learning algorithm to both synthetic and real-world datasets, demonstrating its scalability and versatility. We find that using the structured approximation to the posterior results in models with significantly higher held-out likelihood.}\n}\n\n\n% TOMMI\n\n\n
\n
\n\n\n
\n Gaussian state space models have been used for decades as generative models of sequential data. They admit an intuitive probabilistic interpretation, have a simple functional form, and enjoy widespread adoption. We introduce a unified algorithm to efficiently learn a broad class of linear and non-linear state space models, including variants where the emission and transition distributions are modeled by deep neural networks. Our learning algorithm simultaneously learns a compiled inference network and the generative model, leveraging a structured variational approximation parameterized by recurrent neural networks to mimic the posterior distribution. We apply the learning algorithm to both synthetic and real-world datasets, demonstrating its scalability and versatility. We find that using the structured approximation to the posterior results in models with significantly higher held-out likelihood.\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Aspect-augmented Adversarial Networks for Domain Adaptation.\n \n \n \n \n\n\n \n Zhang, Y.; Barzilay, R.; and Jaakkola, T.\n\n\n \n\n\n\n
Transactions of the Association for Computational Linguistics (TACL). 2017.\n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{Zhang_etal-TACL2017,\ntitle = {Aspect-augmented Adversarial Networks for Domain Adaptation},\nauthor = {Y. Zhang and R. Barzilay and T. Jaakkola},\njournal = {Transactions of the Association for Computational Linguistics (TACL)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/Zhang_etal-TACL17.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n A causal framework for explaining the predictions of black-box sequence-to-sequence models.\n \n \n \n \n\n\n \n Melis, D. A.; and Jaakkola, T.\n\n\n \n\n\n\n In
Empirical Methods in Natural Language Processing (EMNLP), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{AlvJaa_EMNLP2017,\nauthor = {D. Alvarez Melis and T. Jaakkola},\ntitle = {A causal framework for explaining the predictions of black-box sequence-to-sequence models},\nbooktitle = {Empirical Methods in Natural Language Processing (EMNLP)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/AlvJaa_EMNLP2017.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Deriving Neural Architectures from Sequence and Graph Kernels.\n \n \n \n \n\n\n \n Lei, T.; Jin, W.; Barzilay, R.; and Jaakkola, T.\n\n\n \n\n\n\n In
International Conference on Machine Learning (ICML), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{Lei_etal_ICML2017,\nauthor = {T. Lei and W. Jin and R. Barzilay and T. Jaakkola},\ntitle = {Deriving Neural Architectures from Sequence and Graph Kernels},\nbooktitle = {International Conference on Machine Learning (ICML)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/Lei_etal_ICML2017.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Sequence to Better Sequence: Continuous Revision of Combinatorial Structures.\n \n \n \n \n\n\n \n Mueller, J.; Gifford, D.; and Jaakkola, T.\n\n\n \n\n\n\n In
International Conference on Machine Learning (ICML), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{Mueller_etal_ICML2017,\nauthor = {J. Mueller and D. Gifford and T. Jaakkola},\ntitle = {Sequence to Better Sequence: Continuous Revision of Combinatorial Structures},\nbooktitle = {International Conference on Machine Learning (ICML)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/seq2betterseqICML17.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Learning Sleep Stages from Radio Signals: A Conditional Adversarial Architecture.\n \n \n \n \n\n\n \n Zhao, M.; Yue, S.; Katabi, D.; Jaakkola, T.; and Bianchi, M.\n\n\n \n\n\n\n In
International Conference on Machine Learning (ICML), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{Zhao_etal_ICML2017,\nauthor = {M. Zhao and S. Yue and D. Katabi and T. Jaakkola and M. Bianchi},\ntitle = {Learning Sleep Stages from Radio Signals: A Conditional Adversarial Architecture},\nbooktitle = {International Conference on Machine Learning (ICML)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/Zhao_etal_ICML2017.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Modeling Persistent Trends in Distributions.\n \n \n \n \n\n\n \n Mueller, J.; Jaakkola, T.; and Gifford, D.\n\n\n \n\n\n\n
Journal of the American Statistical Association. 2017.\n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{Mueller_etal_JASA_2017,\nauthor = {J. Mueller and T. Jaakkola and D. Gifford},\ntitle = {Modeling Persistent Trends in Distributions},\njournal = {Journal of the American Statistical Association},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/persistentTrends.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Prediction of Organic Reaction Outcomes Using Machine Learning.\n \n \n \n \n\n\n \n Coley, C. W.; Barzilay, R.; Jaakkola, T.; Green, W. H.; and Jensen, K. F.\n\n\n \n\n\n\n
ACS Central Science. 2017.\n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n \n doi\n \n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 3 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{Connor_etal_ACS_2017,\ntitle = {Prediction of Organic Reaction Outcomes Using Machine Learning},\nauthor = {C. W. Coley and R. Barzilay and T. Jaakkola and W. H. Green and K. F. Jensen},\njournal = {ACS Central Science},\nDOI = {DOI: 10.1021/acscentsci.7b00064},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/Connor_etal_ACS_2017.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Tree Structured Decoding with Doubly Recurrent Neural Networks.\n \n \n \n \n\n\n \n Alvarez-Melis, D.; and Jaakkola, T.\n\n\n \n\n\n\n In
International Conference on Learning Representations (ICLR), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{Alvarez_etal_ICLR2017,\nauthor = {D. Alvarez-Melis and T. Jaakkola},\ntitle = {Tree Structured Decoding with Doubly Recurrent Neural Networks},\nbooktitle = {International Conference on Learning Representations (ICLR)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/AlvJaa_ICLR2017.pdf},\n}\n\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Learning Optimal Interventions.\n \n \n \n \n\n\n \n Mueller, J.; Reshef, D.; Du, G.; and Jaakkola, T.\n\n\n \n\n\n\n In
Artificial Intelligence and Statistics (AISTATS), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{Mueller_etal_aistats2017,\nauthor = {J. Mueller and D. Reshef and G. Du and T. Jaakkola},\ntitle = {Learning Optimal Interventions},\nbooktitle = {Artificial Intelligence and Statistics (AISTATS)},\nyear = {2017},\nurl_pdf = {https://people.csail.mit.edu/tommi/papers/Mueller_etal_aistats2017.pdf},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Focused Model-Learning and Planning for Non-Gaussian Continuous State-Action Systems.\n \n \n \n \n\n\n \n Wang, Z.; Jegelka, S.; Kaelbling, L. P.; and Lozano-Perez, T.\n\n\n \n\n\n\n In
IEEE International Conference on Robotics and Automation (ICRA), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{wang17icra,\n author = \t {Zi Wang and Stefanie Jegelka and Leslie Pack Kaelbling and Tomas Lozano-Perez},\n title = \t {Focused Model-Learning and Planning for Non-Gaussian Continuous State-Action Systems},\n booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},\n url_pdf = {http://arxiv.org/abs/1607.07762},\n year = \t 2017}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Multiple wavelength sensing array design.\n \n \n \n\n\n \n Shulkind, G.; Jegelka, S.; and Wornell, G. W.\n\n\n \n\n\n\n In
ICASSP, 2017. \n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{shulkind17,\n author = \t {Gal Shulkind and Stefanie Jegelka and G. W. Wornell},\n title = \t {Multiple wavelength sensing array design},\n booktitle = {ICASSP},\n year = \t 2017}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Deep Metric Learning via Facility Location.\n \n \n \n \n\n\n \n Song, H. O.; Jegelka, S.; Rathod, V.; and Murphy, K.\n\n\n \n\n\n\n In
International Conference on Computer Vision and Pattern Recognition (CVPR), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{song17,\n author = \t {Hyun Oh Song and Stefanie Jegelka and Vivek Rathod and Kevin Murphy},\n title = \t {Deep Metric Learning via Facility Location},\n booktitle = {International Conference on Computer Vision and Pattern Recognition (CVPR)},\n year = \t 2017,\n url_pdf = {http://openaccess.thecvf.com/content_cvpr_2017/papers/Song_Deep_Metric_Learning_CVPR_2017_paper.pdf}}\n\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Robust Budget Allocation via Continuous Submodular Functions.\n \n \n \n \n\n\n \n Staib, M.; and Jegelka, S.\n\n\n \n\n\n\n In
International Conference on Machine Learning (ICML), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 1 download\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{staib17,\n author = \t {Matthew Staib and Stefanie Jegelka},\n title = \t {Robust Budget Allocation via Continuous Submodular Functions},\n booktitle = {International Conference on Machine Learning (ICML)},\n year = \t 2017,\n url_pdf = {http://proceedings.mlr.press/v70/staib17a/staib17a.pdf}}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Max-value entropy search for efficient Bayesian Optimization.\n \n \n \n \n\n\n \n Wang, Z.; and Jegelka, S.\n\n\n \n\n\n\n In
International Conference on Machine Learning (ICML), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 2 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{wang17mes,\n author = \t {Zi Wang and Stefanie Jegelka},\n title = \t {Max-value entropy search for efficient Bayesian Optimization},\n booktitle = {International Conference on Machine Learning (ICML)},\n year = \t 2017,\n url_pdf = {http://proceedings.mlr.press/v70/wang17e/wang17e.pdf}}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Batched High-dimensional Bayesian Optimization via Structural Kernel Learning.\n \n \n \n \n\n\n \n Wang, Z.; Li, C.; Jegelka, S.; and Kohli, P.\n\n\n \n\n\n\n In
International Conference on Machine Learning (ICML), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{wangLi17,\n author = \t {Zi Wang and Chengtao Li and Stefanie Jegelka and Pushmeet Kohli},\n title = \t {Batched High-dimensional Bayesian Optimization via Structural Kernel Learning.},\n booktitle = {International Conference on Machine Learning (ICML)},\n year = \t 2017,\n url_pdf = {http://proceedings.mlr.press/v70/wang17h/wang17h.pdf}}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Wasserstein k-means++ for Cloud Regime Histogram Clustering.\n \n \n \n\n\n \n Staib, M.; and Jegelka, S.\n\n\n \n\n\n\n In
Climate Informatics, 2017. \n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{staib17ci,\n author = \t {Matthew Staib and Stefanie Jegelka},\n title = \t {Wasserstein k-means++ for Cloud Regime Histogram Clustering},\n booktitle = {Climate Informatics},\n year = \t 2017}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Parallel Streaming Wasserstein Barycenters.\n \n \n \n \n\n\n \n Staib, M.; Claici, S.; Solomon, J.; and Jegelka, S.\n\n\n \n\n\n\n In
Advances in Neural Information Processing Systems (NIPS), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{staibCSJ17,\n author = \t {Matthew Staib and\n Sebastian Claici and\n Justin Solomon and\n Stefanie Jegelka},\n title = \t {Parallel Streaming {W}asserstein Barycenters},\n booktitle = {Advances in Neural Information Processing Systems (NIPS)},\n url_pdf = {https://arxiv.org/abs/1705.07443},\n year = \t 2017}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Polynomial Time Algorithms for Dual Volume Sampling.\n \n \n \n \n\n\n \n Li, C.; Jegelka, S.; and Sra, S.\n\n\n \n\n\n\n In
Advances in Neural Information Processing Systems (NIPS), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{li17dual,\n author = \t {Chengtao Li and Stefanie Jegelka and Suvrit Sra},\n title = \t {Polynomial Time Algorithms for Dual Volume Sampling},\n booktitle = {Advances in Neural Information Processing Systems (NIPS)},\n url_pdf = {https://arxiv.org/abs/1703.02674},\n year = \t 2017}\n\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Virtual screening of inorganic materials synthesis parameters with deep learning.\n \n \n \n \n\n\n \n Kim, E.; Huang, K.; Jegelka, S.; and Olivetti, E.\n\n\n \n\n\n\n
npj Computational Materials, 3(53). 2017.\n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@Article{kim17,\n author = \t {Edward Kim and Kevin Huang and Stefanie Jegelka and Elsa Olivetti},\n title = \t {Virtual screening of inorganic materials synthesis parameters with deep learning},\n journal = \t {npj Computational Materials},\n year = \t 2017,\n volume = \t 3,\n number = \t 53,\n url_pdf = {https://www.nature.com/articles/s41524-017-0055-6}}\n\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Logarithmic inequalities under a symmetric polynomial dominance order.\n \n \n \n\n\n \n Sra, S.\n\n\n \n\n\n\n
Proceedings American Mathematical Society (PAMS). Oct 2017.\n
ıt Accepted.\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@Article{sra.esym,\n author = {Suvrit Sra},\n title = {Logarithmic inequalities under a symmetric polynomial dominance order},\n journal = {Proceedings American Mathematical Society (PAMS)},\n year = 2017,\n month = {Oct},\n note = {{\\it Accepted.}},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Elementary symmetric polynomials for optimal experimental design.\n \n \n \n \n\n\n \n Mariet, Z.; and Sra, S.\n\n\n \n\n\n\n In
Advances in Neural Information Processing Systems (NIPS), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{marietSra17b,\n author = \t {Zelda Mariet and Suvrit Sra},\n title = \t {Elementary symmetric polynomials for optimal experimental design},\n booktitle = {Advances in Neural Information Processing Systems (NIPS)},\n url_pdf = {https://arxiv.org/abs/1705.09677},\n year = \t 2017}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Combinatorial topic modling using small variance asymptotics.\n \n \n \n \n\n\n \n Jiang, K.; Sra, S.; and Kulis, B.\n\n\n \n\n\n\n In
Artificial Intelligence and Statistics (AISTATS), 2017. \n
\n\n
\n\n
\n\n
\n\n \n \n
pdf\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{sraKulis, \nauthor = {Ke Jiang and Suvrit Sra and Brian Kulis},\ntitle = {Combinatorial topic modling using small variance asymptotics},\nbooktitle = {Artificial Intelligence and Statistics (AISTATS)},\nyear = {2017},\nurl_pdf = {http://arxiv.org/abs/1604.02027}\n}\n\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Size-Independent Sample Complexity of Neural Networks.\n \n \n \n\n\n \n Golowich, N.; Rakhlin, A.; and Shamir, O.\n\n\n \n\n\n\n
arXiv preprint arXiv:1712.06541. 2017.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{golowich2017size,\n title={Size-Independent Sample Complexity of Neural Networks},\n author={Golowich, Noah and Rakhlin, Alexander and Shamir, Ohad},\n journal={arXiv preprint arXiv:1712.06541},\n year={2017}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Fisher-rao metric, geometry, and complexity of neural networks.\n \n \n \n\n\n \n Liang, T.; Poggio, T.; Rakhlin, A.; and Stokes, J.\n\n\n \n\n\n\n
arXiv preprint arXiv:1711.01530. 2017.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{liang2017fisher,\n title={Fisher-rao metric, geometry, and complexity of neural networks},\n author={Liang, Tengyuan and Poggio, Tomaso and Rakhlin, Alexander and Stokes, James},\n journal={arXiv preprint arXiv:1711.01530},\n year={2017}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Non-convex learning via Stochastic Gradient Langevin Dynamics: a nonasymptotic analysis.\n \n \n \n\n\n \n Raginsky, M.; Rakhlin, A.; and Telgarsky, M.\n\n\n \n\n\n\n In
COLT, 2017. \n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{RagRakTel17,\n title = {Non-convex learning via Stochastic Gradient Langevin Dynamics: a nonasymptotic analysis},\n author = {Maxim Raginsky and Alexander Rakhlin and Matus Telgarsky},\n booktitle = {COLT},\n year = {2017},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n ZigZag: A New Approach to Adaptive Online Learning.\n \n \n \n\n\n \n Foster, D. J.; Rakhlin, A.; and Sridharan, K.\n\n\n \n\n\n\n In
COLT, 2017. \n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{FosRakSri17a,\n title = {ZigZag: A New Approach to Adaptive Online Learning},\n author = {Dylan J. Foster and Alexander Rakhlin and Karthik Sridharan},\n booktitle = {COLT},\n year = {2017},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n On Equivalence of Martingale Tail Bounds and Deterministic Regret Inequalities.\n \n \n \n\n\n \n Rakhlin, A.; and Sridharan, K.\n\n\n \n\n\n\n In
COLT, 2017. \n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@InProceedings{RakSri17a,\n title = {On Equivalence of Martingale Tail Bounds and Deterministic Regret Inequalities},\n author = {Alexander Rakhlin and Karthik Sridharan},\n booktitle = {COLT},\n year = {2017},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Efficient Online Multiclass Prediction on Graphs via Surrogate Losses.\n \n \n \n\n\n \n Rakhlin, A.; and Sridharan, K.\n\n\n \n\n\n\n In
Artificial Intelligence and Statistics, 2017. \n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@inproceedings{rakhlin2017efficient,\n title={Efficient Online Multiclass Prediction on Graphs via Surrogate Losses},\n author={Rakhlin, Alexander and Sridharan, Karthik},\n booktitle={Artificial Intelligence and Statistics},\n year={2017}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n On detection and structural reconstruction of small-world random networks.\n \n \n \n\n\n \n Cai, T.; Liang, T.; and Rakhlin, A.\n\n\n \n\n\n\n
IEEE Transactions on Network Science and Engineering, 4(3): 165–176. 2017.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{cai2017detection,\n title={On detection and structural reconstruction of small-world random networks},\n author={Cai, Tony and Liang, Tengyuan and Rakhlin, Alexander},\n journal={IEEE Transactions on Network Science and Engineering},\n volume={4},\n number={3},\n pages={165--176},\n year={2017},\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Computational and statistical boundaries for submatrix localization in a large noisy matrix.\n \n \n \n\n\n \n Cai, T T.; Liang, T.; Rakhlin, A.; and others\n\n\n \n\n\n\n
The Annals of Statistics, 45(4): 1403–1430. 2017.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@article{cai2017computational,\n title={Computational and statistical boundaries for submatrix localization in a large noisy matrix},\n author={Cai, T Tony and Liang, Tengyuan and Rakhlin, Alexander and others},\n journal={The Annals of Statistics},\n volume={45},\n number={4},\n pages={1403--1430},\n year={2017},\n}\n\n
\n
\n\n\n\n
\n\n\n\n\n\n