Willkommen!
Ryota Tomioka, PhD
 I was a member of
Intelligent Data Analysis Group at Fraunhofer FIRST and TU Berlin (Berlin, Germany) from Nov 2005 to Oct 2007 ... to get this incredible poster!
 I took part in Machine Learning Summer School 2007 in Tübingen, Germany.
 2008/03/24 I got my PhD (Thesis advisor: Prof. Kazuyuki Aihara) with "Univ. of Tokyo, School of Information Science and Technology President's Award".
 2008/04/01 I moved to Masashi Sugiyama's group at Tokyo Tech.
 2009/04/01 I moved back to Department of Mathematical Informatics, University of Tokyo to join Kenji Yamanishi's group.
 2009/12/12 Check out my talk at the Optimization workshop at NIPS.
 2010/6/22: Slides and code for my talk at ICML 2010.
 2011/8/26: Slides and exercises for my lecture at DTU Ph.D. Summer Course on Machine Learning.
 2012/1/26: Slides for my talk in informatics seminar at Kyoto University.
 2012/8/15: Slides and exercises for my lecture at DTU Ph.D. Summer Course on Advanced Machine Learning. This year, I have included an interactive demo to play with convex conjugate function pairs in MATLAB.
 2012/9 and 2013/3: I had a great time at Mathematisches Forschungsinstitut Oberwolfach with Leibnitz fellow Franz Király. [Pic]
 2013/8/14: Slides and other materials for my lecture at DTU Machine Learning Summer School.
 2013/10: I joined TTIC as a Research Assistant Professor.
 2014/3: I gave intived talks at ISM and Kyoto University.
 2014/8: Slides and other materials for my lecture at DTU Machine Learning Summer School.
 2015/9: I gave a lecture at MLSS 2015 Kyoto on tensor decompositions. [Slides and iPython Notebooks]
 2015/11: After spending two wonderful years in Chicago, I joined MSR Cambridge as a Researcher.
Research Interests
 Learning algorithms for atomic simulations (New!)
 Previous interests:
Publications
Book chapters

Augmented Lagrangian Methods for Learning, Selecting, and Combining Features. R. Tomioka and T. Suzuki and M. Sugiyama. In Suvrit Sra, Sebastian Nowozin, and Stephen J. Wright Eds. Optimization for Machine Learning, MIT Press, 2011. [Amazon.com]

Lowrank tensor denoising and recovery via convex optimization. R. Tomioka, T. Suzuki, K. Hayashi, and H. Kashima. In J. Suykens, M. Signoretto, & A. Argyriou (editors), Regularization, Optimization, Kernels, and Support Vector Machines, 2014. [Amazon.com]
Refereed conference papers

An Informationtheoretic Approach to Distribution Shifts. Marco Federici, Ryota Tomioka, Patrick Forré. NeurIPS 2021.

Regularized Policies are Reward Robust. Hisham Husain, Kamil Ciosek, Ryota Tomioka. AISTATS 2021.

On the Loss Landscape of Adversarial Training: Identifying Challenges and How to Overcome Them. Chen Liu, Mathieu Salzmann, Tao Lin, Ryota Tomioka, Sabine Süsstrunk. NeurIPS 2020.

Conservative Uncertainty Estimation By Fitting Prior Networks. Kamil Ciosek, Vincent Fortuin, Ryota Tomioka, Katja Hofmann, Richard Turner. ICLR 2020.

Continuous Hierarchical Representations with Poincaré Variational AutoEncoders. Emile Mathieu, Charline Le Lan, Chris J. Maddison, Ryota Tomioka, Yee Whye Teh. NeurIPS 2019.

On Certifying NonUniform Bounds against Adversarial Attacks. Chen Liu, Ryota Tomioka, Volkan Cevher. ICML 2019.

Multilevel variational autoencoder: Learning disentangled representations from grouped observations. Diane Bouchacourt, Ryota Tomioka, Sebastian Nowozin. AAAI 2018.

QSGD: CommunicationEfficient SGD via Gradient Quantization and Encoding. Dan Alistarh, Demjan Grubic, Jerry Li, Ryota Tomioka, Milan Vojnovic. NIPS, 2017. [Slides]

fGAN: Training Generative Neural Samplers using Variational Divergence Minimization. Sebastian Nowozin, Botond Cseke, Ryota Tomioka. NIPS, 2016.

DataDependent Path Normalization in Neural Networks.
Behnam Neyshabur, Ryota Tomioka, Ruslan Salakhutdinov, Nathan Srebro. ICLR, 2016.

Interpolating Convex and NonConvex Tensor Decompositions via the Subspace Norm. Qinqing Zheng, Ryota Tomioka. In Advances in NIPS 28, pages 31063113, 2015.

NormBased Capacity Control in Neural Networks. Behnam Neyshabur, Ryota Tomioka, and Nathan Srebro. COLT, 2015.

In Search of the Real Inductive Bias: On the Role of Implicit Regularization in Deep Learning.
Behnam Neyshabur, Ryota Tomioka, and Nathan Srebro. Workshop track poster at ICLR, 2015.

Multitask learning meets tensor factorization: task imputation via convex optimization. K. Wimalawarne, M. Sugiyama, and R. Tomioka. In Advances in NIPS 27, pages 28252833, 2014.

Convex Tensor Decomposition via Structured Schatten Norm Regularization. Ryota Tomioka and Taiji Suzuki. In NIPS 2013. [Poster] [Preliminary version]

Nonnegative Multiple Tensor Factorization. Koh Takeuchi, RyotaTomioka, Katsuhiko Ishiguro, Akisato Kimura, and Hiroshi Sawada. In Proc. ICDM 2013.

Quantitative Prediction of Glaucomatous Visual Field Loss from Few Measurements. ZengHan Liang, Ryota Tomioka, Hiroshi Murata, Ryo Asaoka, and Kenji Yamanishi. In Proc. ICDM 2013.

Infinite Positive Semidefinite Tensor Factorization with Application to Music Signal Analysis. Kazuyoshi Yoshii, Ryota Tomioka, Daichi Mochihashi, and Masataka Goto. In Proc. 30th International Conference on Machine Learning (ICML 2013). 2013, Atlanta, USA. [SI & Demo audio].

Perfect Dimensionality Recovery by Variational Bayesian PCA. Shinichi Nakajima, Ryota Tomioka, Masashi Sugiyama, S. Derin Babacan. In Advances in NIPS 25. 2012, Lake Tahoe, NV, USA.

A Combinatorial Algebraic Approach for the Identifiability of LowRank Matrix Completion. Franz Király and Ryota Tomioka. In Proc. of the 29th International Conference on Machine Learning (ICML 2012), Edinburgh, UK.

A Bayesian Analysis of the Radioactive Releases of Fukushima. Ryota Tomioka and Morten Mørup. AI & Statistics 2012, La Palma, Spain. (JMLR W&CP 22: 12431251) [Software]

Statistical Performance of Convex Tensor Decomposition. Ryota Tomioka, Taiji Suzuki, Kohei Hayashi, and Hisashi Kashima. Advances in NIPS 24. 2011, Granada, Spain. [Poster]

Discovering Emerging Topics in Social Streams via Link Anomaly Detection. Toshimitsu Takahashi, Ryota Tomioka, and Kenji Yamanishi. ICDM 2011 IEEE International Conference on Data Mining, 2011, Vancouver, Canada. [Longer version]

Tensor Factorization Using Auxiliary Information. Atsuhiro Narita, Kohei Hayashi, Ryota Tomioka and Hisashi Kashima. In Proc. European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), pp. 501516, Athens, Greece, 2011. Best Student Paper Award in Data Mining.

Global Analytic Solution for Variational Bayesian Matrix Factorization. S. Nakajima, M. Sugiyama, R. Tomioka, In Advances in NIPS 23, pp. 1768–1776. 2010, Vancouver, Canada.

A Fast Augmented Lagrangian Algorithm for Learning LowRank Matrices. Ryota Tomioka, Taiji Suzuki, Masashi Sugiyama, and Hisashi Kashima, Proc. of the 27 th Annual International Conference on Machine Learning (ICML 2010), Haifa, Israel, 2010. [Slides] [Software]

Invariant Common Spatial Patterns: Alleviating Nonstationarities in BrainComputer Interfacing, B. Blankertz, M. Kawanabe, R. Tomioka, F. Hohlefeld, V. Nikulin, and K.R. Müller, NIPS 2007, Vancouver, Canada, Dec. 2007. [Bibtex] [Spotlight]

Classifying Matrices with a Spectral Regularization, Ryota Tomioka and Kazuyuki Aihara, Proc. of the 24th Annual International Conference on Machine Learning (ICML2007), pp. 895902, ACM Press. Oregon, USA, June, 2007. [BibTex][Slides][Software]

Logistic Regression for Single Trial EEG Classification, Ryota Tomioka, Kazuyuki Aihara, and KlausRobert Müller, Advances in Neural Information Processing Systems 19, pp. 13771384, MIT Press. NIPS 2006, Vancouver, Canada, Dec. 2006. [BibTex] [Poster] [Software]

Optimizing Spectral Filters for Single Trial EEG Classification, R. Tomioka, G. Dornhege, G. Nolte, K. Aihara and K.R. Müller, LNCS 4174, pp. 414423, Springer. DAGM2006, Berlin, Germany, Sep. 2006. [Springer]
Refereed journal papers

Theoretical and Experimental Analyses of TensorBased Regression and Classification.
Kishan Wimalawarne, Ryota Tomioka, Masashi Sugiyama, Neural Computation, 28 (4), 686715, 2016.

Condition for Perfect Dimensionality Recovery by Variational Bayesian PCA. S. Nakajima, R. Tomioka, M. Sugiyama, S. D. Babacan. Journal of Machine Learning Research 16, 37573811, 2015.

The Algebraic Combinatorial Approach for LowRank Matrix Completion. Franz J. Király, Louis Theran, and Ryota Tomioka. Journal of Machine Learning Research 16, 13911436, 2015.

Discovering Emerging Topics in Social Streams via Link Anomaly Detection. Toshimitsu Takahashi, Ryota Tomioka, Kenji Yamanishi, IEEE Transactions on Knowledge and Data Engineering, 26 (1), pp. 120130, 2014.

Global Analytic Solution of Fullyobserved Variational Bayesian Matrix Factorization. Shinichi Nakajima, Masashi Sugiyama, S. Derin Babacan, Ryota Tomioka, Journal of Machine Learning Research, 14:137, 2013.

Tensor factorization using auxiliary information. Atsuhiro Narita, Kohei Hayashi, Ryota Tomioka, Hisashi Kashima, Data Mining and Knowledge Discovery, 25(2), pp. 298324, 2012.

SpicyMKL: A Fast Algorithm for Multiple Kernel Learning with Thousands of Kernels. Taiji Suzuki and Ryota Tomioka, Machine Learning, 2011. Accepted. [Preprint] [Software]

SuperLinear Convergence of Dual Augmented Lagrangian Algorithm for Sparse Learning. Ryota Tomioka, Taiji Suzuki, and Masashi Sugiyama. Journal of Machine Learning Research, 12(May):15371586, 2011. A preliminary version was presented at NIPS2009 Workshop: Optimization for Machine Learning [Slides] [Video] [Software]

LargeScale EEG/MEG Source Localization with Spatial Flexibility. S. Haufe, R. Tomioka, T. Dickhaus, C. Sannelli, B. Blankertz, G. Nolte, K.R. Müller, Neuroimage, 2010. Accepted.

Modeling sparse connectivity between underlying brain sources for EEG/MEG. Stefan Haufe, Ryota Tomioka, Guido Nolte, KlausRobert Müller, and Motoaki Kawanabe, IEEE Trans. Biomed. Eng. 57(8), pp. 19541963, 2010.

A regularized discriminative framework for EEG analysis with application to braincomputer interface. Ryota Tomioka and KlausRobert Müller, Neuroimage, 49 (1) pp. 415432, 2010. [Software]

Dual Augmented Lagrangian Method for Efficient Sparse Reconstruction, Ryota Tomioka and Masashi Sugiyama, IEEE Signal Proccesing Letters, 16 (12) pp. 10671070, 2009. [Preprint] [Software]

Optimizing Spatial Filters for Robust EEG SingleTrial Analysis, Benjamin Blankertz, Ryota Tomioka, Steven Lemm, Motoaki Kawanabe, and KlausRobert Müller, IEEE Signal Proc. Magazine, 25 (1), pp. 4156, 2008.

Multivariate Analysis of Noise in Genetic Regulatory Networks, Ryota TOMIOKA, Hidenori KIMURA, Tetsuya J. KOBAYASHI and Kazuyuki AIHARA, Journal of Theoretical Biology, 229 (4), 501521, 21 August 2004. (preprint)
Other talks

On the extension of trace norm to tensors. Ryota Tomioka, Kohei Hayashi, and Hisashi Kashima, NIPS2010 Workshop: Tensors, kernels and machine learning, Whistler, Canada, 2010. [Slides] [Software]

Sparse learning with duality gap guarantee, Ryota Tomioka, Masashi Sugiyama. In NIPS workshop OPT 2008 Optimization for Machine Learning, 2008.

Combined classification and channel/basis selection with L1L2 regularization with application to P300 speller system, Ryota Tomioka and Stefan Haufe, Proc. 4th International BCI Workshop and Training Course 2008, Graz, Austria, Sep. 2008. [Slides]

Adapting Spatial Filtering Methods for Nonstationary BCIs, Ryota Tomioka, Jeremy Hill, Benjamin Blankertz, and Kazuyuki Aihara, Proc. IBIS2006, pp.6570. IBIS2006, Osaka, Japan, Nov. 2006.

An Iterative Algorithm for SpatioTemporal Filter Optimization, R. Tomioka, G. Dornhege, K. Aihara and K.R. Müller, Verlag der Technischen Universität Graz. Proc. 3rd International BCI Workshop and Training Course 2006, Graz, Austria, Sep. 2006.
Others

Gaussian Attention Model and Its Application to Knowledge Base Embedding and Question Answering. Liwen Zhang, John Winn, Ryota Tomioka.

Jointly Learning Multiple Perceptual Similarities. Liwen Zhang, Subhransu Maji, and Ryota Tomioka. Technical report, 2015.

Spectral norm of random tensors. Ryota Tomioka, Taiji Suzuki. Technical report, 2014.

Estimation of lowrank tensors via convex optimization. R. Tomioka and K. Hayashi and H. Kashima. Submitted, 2011. [Slides] [Software]

Regularization Strategies and Empirical Bayesian Learning for MKL. Ryota Tomioka and Taiji Suzuki. A preliminary version was presented at NIPS2010 Workshop: New Directions in Multiple Kernel Learning. [Slides] [Video]

Spectrally Weighted Common Spatial Pattern Algorithm for Single Trial EEG Classification, Ryota Tomioka, Guido Dornhege, Guido Nolte, Benjamin Blankertz, Kazuyuki Aihara, and KlausRobert Müller, Mathematical Engineering Technical Reports (METR200640), July 2006.

Modeling fluctuations in genetic regulatory networks : a structural approach, Ryota TOMIOKA, Hidenori KIMURA, Tetsuya J. KOBAYASHI, and Kazuyuki AIHARA, IEICE technical report. Neurocomputing, 104 (226) pp. 1317.

A GraphBased Analysis of Stochasticity in Intracellular Networks, Tetsuya J. KOBAYASHI, Ryota TOMIOKA and Kazuyuki AIHARA, Mathematical Engineering Technical Reports (METR) 200424, , May 2004.

スパース性に基づく機械学習 (機械学習プロフェッショナルシリーズ). 冨岡 亮太. 講談社, 2015. [サポートページ]
Professional Activities
 Action Editor for Journal of Machine Learing Reseach (2017)
 Editor for Neural Networks (20142016)
 Area Chair for NIPS 2015, 2017, NeurIPS 2018 and ICML 2017

Reviewer for NIPS 2007, 2008, 2010, 2011, 2012, 2013, 2014, 2016ICML 2009, 2013, 2014, 2016. AISTATS 2014, ACML 2009, 2010, 2011, 2012, 2013, 2014, ECML 2010 JMLR, Neural Networks, Neuroimage, Pattern Recognition, IEEE TNN, IEEE TSP, IEEE TBME, IEEE TNSRE, IEEE SPL, IEEE TPAMI

PC member of NIPS 2009 Workshop on Connectivity Inference in Neuroimaging (CINI 2009) and NIPS workshop Optimization for Machine Learning 2011, 2012.
Misc.