On the Learning Dynamics of Two-layer Linear Networks with Label Noise SGD.

Taiji Suzuki http://www.is.titech.ac.jp/~s-taiji/ https://scholar.google.com/citations?user=x8osrBsAAAAJ https://openreview.net/profile?id=~Taiji_Suzuki1 https://orcid.org/0000-0003-3459-1016 Tokyo Institute of Technology, Department of Mathematical and Computing Sciences Tongcheng Zhang Zhanpeng Zhou Mingze Wang Andi Han Wei Huang 0034 Taiji Suzuki Junchi Yan On the Learning Dynamics of Two-layer Linear Networks with Label Noise SGD. 28400-28408 2026 AAAI https://doi.org/10.1609/aaai.v40i33.40069 conf/aaai/2026 db/conf/aaai/aaai2026.html#ZhangZWHHSY26

Yakun Wang 0005 Leyang Wang Song Liu 0002 Taiji Suzuki Zero-Flow Encoders. 2026 February abs/2602.00797 CoRR https://doi.org/10.48550/arXiv.2602.00797 db/journals/corr/corr2602.html#abs-2602-00797 streams/journals/corr

Akifumi Wachi Hirota Kinoshita Shokichi Takakura Rei Higuchi Taiji Suzuki A Relative-Budget Theory for Reinforcement Learning with Verifiable Rewards in Large Language Model Reasoning. 2026 February abs/2602.01523 CoRR https://doi.org/10.48550/arXiv.2602.01523 db/journals/corr/corr2602.html#abs-2602-01523 streams/journals/corr

Shokichi Takakura Akifumi Wachi Rei Higuchi Kohei Miyaguchi Taiji Suzuki Inference-Aware Meta-Alignment of LLMs via Non-Linear GRPO. 2026 February abs/2602.01603 CoRR https://doi.org/10.48550/arXiv.2602.01603 db/journals/corr/corr2602.html#abs-2602-01603 streams/journals/corr

Ryotaro Kawata Taiji Suzuki Transformers as Measure-Theoretic Associative Memory: A Statistical Perspective and Minimax Optimality. 2026 February abs/2602.01863 CoRR https://doi.org/10.48550/arXiv.2602.01863 db/journals/corr/corr2602.html#abs-2602-01863 streams/journals/corr

Hiroki Naganuma Taiji Suzuki Rio Yokota Masahiro Nomura Kohta Ishikawa Ikuro Sato Takeuchi's Information Criteria as Generalization Measures for DNNs Close to NTK Regime. 2026 February abs/2602.23219 CoRR https://doi.org/10.48550/arXiv.2602.23219 db/journals/corr/corr2602.html#abs-2602-23219 streams/journals/corr

Tongcheng Zhang Zhanpeng Zhou Mingze Wang Andi Han Wei Huang 0034 Taiji Suzuki Junchi Yan On the Learning Dynamics of Two-layer Linear Networks with Label Noise SGD. 2026 March abs/2603.10397 CoRR https://doi.org/10.48550/arXiv.2603.10397 db/journals/corr/corr2603.html#abs-2603-10397 streams/journals/corr

Dake Bu Wei Huang 0034 Andi Han Hau-San Wong Qingfu Zhang 0001 Taiji Suzuki Atsushi Nitanda DPRM: A Plug-in Doob h transform-induced Token-Ordering Module for Diffusion Language Models. 2026 April abs/2604.24357 CoRR https://doi.org/10.48550/arXiv.2604.24357 db/journals/corr/corr2604.html#abs-2604-24357 streams/journals/corr

Ryoya Awano Taiji Suzuki The Mechanism of Weak-to-Strong Generalization: Feature Elicitation from Latent Knowledge. 2026 May abs/2605.12908 CoRR https://doi.org/10.48550/arXiv.2605.12908 db/journals/corr/corr2605.html#abs-2605-12908 streams/journals/corr

Guoji Fu Taiji Suzuki Wee Sun Lee Atsushi Nitanda Intrinsic Wasserstein Rates for Score-Based Generative Models on Smooth Manifolds. 2026 May abs/2605.15822 CoRR https://doi.org/10.48550/arXiv.2605.15822 db/journals/corr/corr2605.html#abs-2605-15822 streams/journals/corr

Wei Huang 0034 Andi Han Mingyuan Bai Huanjian Zhou Qixin Zhang Taiji Suzuki Kenji Fukumizu Provably Learning Diffusion Models under the Manifold Hypothesis: Collapse and Refine. 2026 May abs/2605.20235 CoRR https://doi.org/10.48550/arXiv.2605.20235 db/journals/corr/corr2605.html#abs-2605-20235 streams/journals/corr

Rei Higuchi Ryotaro Kawata Akifumi Wachi Shokichi Takakura Kohei Miyaguchi Taiji Suzuki How Neural Reward Models Learn Features for Policy Optimization: A Single-Index Analysis. 2026 May abs/2605.24749 CoRR https://doi.org/10.48550/arXiv.2605.24749 db/journals/corr/corr2605.html#abs-2605-24749 streams/journals/corr

Yuanhe Zhang Yuekai Sun Taiji Suzuki Jason D. Lee Fanghui Liu 0001 LeanMarathon: Toward Reliable AI Co-Mathematicians through Long-Horizon Lean Autoformalization. 2026 June abs/2606.05400 CoRR https://doi.org/10.48550/arXiv.2606.05400 db/journals/corr/corr2606.html#abs-2606-05400 streams/journals/corr

Kishan Wimalawarne Taiji Suzuki Sophie Langer Learning Green's function efficiently using low-rank approximations. 214 2025 October 114 Mach. Learn. 10 https://doi.org/10.1007/s10994-025-06845-7 db/journals/ml/ml114.html#WimalawarneSL25 streams/journals/ml

Tomoya Murata Atsushi Nitanda Taiji Suzuki Clustered Invariant Risk Minimization. 1612-1620 2025 AISTATS https://proceedings.mlr.press/v258/murata25a.html conf/aistats/2025 db/conf/aistats/aistats2025.html#MurataNS25 streams/conf/aistats Wei Huang 0034 Yuan Cao 0006 Haonan Wang Xin Cao 0001 Taiji Suzuki Quantifying the Optimization and Generalization Advantages of Graph Neural Networks Over Multilayer Perceptrons. 2854-2862 2025 AISTATS https://proceedings.mlr.press/v258/huang25a.html conf/aistats/2025 db/conf/aistats/aistats2025.html#HuangCWCS25 streams/conf/aistats Wei Huang 0034 Wuyang Chen 0001 Zhiqiang Xu 0003 Zhangyang Wang Taiji Suzuki Exact and Rich Feature Learning Dynamics of Two-Layer Linear Networks. 1087-1111 2025 CPAL https://proceedings.mlr.press/v280/huang25a.html conf/cpal/2025 db/conf/cpal/cpal2025.html#HuangCXWS25 streams/conf/cpal Kenji Fukumizu Taiji Suzuki Noboru Isobe Kazusato Oko Masanori Koyama Flow matching achieves almost minimax optimal convergence. 2025 ICLR https://openreview.net/forum?id=2OMyAFjiJJ conf/iclr/2025 db/conf/iclr/iclr2025.html#FukumizuSIOK25 Ryotaro Kawata Kazusato Oko Atsushi Nitanda Taiji Suzuki Direct Distributional Optimization for Provable Alignment of Diffusion Models. 2025 ICLR https://openreview.net/forum?id=Nvw2szDdmI conf/iclr/2025 db/conf/iclr/iclr2025.html#KawataONS25 Juno Kim Dimitri Meunier Arthur Gretton Taiji Suzuki Zhu Li Optimality and Adaptivity of Deep Neural Features for Instrumental Variable Regression. 2025 ICLR https://openreview.net/forum?id=ReItdfwMcg conf/iclr/2025 db/conf/iclr/iclr2025.html#KimMGSL25 Juno Kim Taiji Suzuki Transformers Provably Solve Parity Efficiently with Chain of Thought. 2025 ICLR https://openreview.net/forum?id=n2NidsYDop conf/iclr/2025 db/conf/iclr/iclr2025.html#KimS25 Bingrui Li Wei Huang 0034 Andi Han Zhanpeng Zhou Taiji Suzuki Jun Zhu 0001 Jianfei Chen 0001 On the Optimization and Generalization of Two-layer Transformers with Sign Gradient Descent. 2025 ICLR https://openreview.net/forum?id=97rOQDPmk2 conf/iclr/2025 db/conf/iclr/iclr2025.html#LiHHZSZC25 Naoki Nishikawa Taiji Suzuki State Space Models are Provably Comparable to Transformers in Dynamic Token Selection. 2025 ICLR https://openreview.net/forum?id=QFgbJOYJSE conf/iclr/2025 db/conf/iclr/iclr2025.html#NishikawaS25 Toshimitsu Uesaka Taiji Suzuki Yuhta Takida Chieh-Hsin Lai Naoki Murata Yuki Mitsufuji Weighted Point Set Embedding for Multimodal Contrastive Learning Toward Optimal Similarity Metric. 2025 ICLR https://openreview.net/forum?id=uSz2K30RRd conf/iclr/2025 db/conf/iclr/iclr2025.html#UesakaSTLMM25 Dake Bu Wei Huang 0034 Andi Han Atsushi Nitanda Qingfu Zhang 0001 Hau-San Wong Taiji Suzuki Provable In-Context Vector Arithmetic via Retrieving Task Concepts. 2025 ICML https://proceedings.mlr.press/v267/bu25a.html https://openreview.net/forum?id=DbUmeNnNpt conf/icml/2025 db/conf/icml/icml2025.html#Bu0HN0WS25 Andi Han Wei Huang 0034 Zhanpeng Zhou Gang Niu 0001 Wuyang Chen 0001 Junchi Yan Akiko Takeda Taiji Suzuki On the Role of Label Noise in the Feature Learning Process. 2025 ICML https://proceedings.mlr.press/v267/han25c.html https://openreview.net/forum?id=kwHvs1UdTM conf/icml/2025 db/conf/icml/icml2025.html#Han0Z00YTS25 Rei Higuchi Taiji Suzuki Direct Density Ratio Optimization: A Statistically Consistent Approach to Aligning Large Language Models. 2025 ICML https://proceedings.mlr.press/v267/higuchi25a.html https://openreview.net/forum?id=jDdaysR6bz conf/icml/2025 db/conf/icml/icml2025.html#HiguchiS25 Ryotaro Kawata Kohsei Matsutani Yuri Kinoshita Naoki Nishikawa Taiji Suzuki Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning. 2025 ICML https://proceedings.mlr.press/v267/kawata25a.html https://openreview.net/forum?id=2xmOEbpYv1 conf/icml/2025 db/conf/icml/icml2025.html#KawataMKNS25 Juno Kim Denny Wu Jason D. Lee Taiji Suzuki Metastable Dynamics of Chain-of-Thought Reasoning: Provable Benefits of Search, RL and Distillation. 2025 ICML https://proceedings.mlr.press/v267/kim25aj.html https://openreview.net/forum?id=2HJcVtuovs conf/icml/2025 db/conf/icml/icml2025.html#KimWLS25 Naoki Nishikawa Yujin Song Kazusato Oko Denny Wu Taiji Suzuki Nonlinear transformers can perform inference-time feature learning. 2025 ICML https://proceedings.mlr.press/v267/nishikawa25a.html https://openreview.net/forum?id=xQTSvP57C3 conf/icml/2025 db/conf/icml/icml2025.html#NishikawaSOWS25 Atsushi Nitanda Anzelle Lee Damian Tan Xing Kai Mizuki Sakaguchi Taiji Suzuki Propagation of Chaos for Mean-Field Langevin Dynamics and its Application to Model Ensemble. 2025 ICML https://proceedings.mlr.press/v267/nitanda25a.html https://openreview.net/forum?id=SjT6JK4KZv conf/icml/2025 db/conf/icml/icml2025.html#NitandaLKSS25 Rom N. Parnichkun Neehal Tumma Armin W. Thomas Alessandro Moro Qi An 0001 Taiji Suzuki Atsushi Yamashita Michael Poli Stefano Massaroli Quantifying Memory Utilization with Effective State-Size. 2025 ICML https://proceedings.mlr.press/v267/parnichkun25a.html https://openreview.net/forum?id=YKAnIvY5hf conf/icml/2025 db/conf/icml/icml2025.html#ParnichkunTTMAS25 Wei Huang 0034 Andi Han Yujin Song Yilan Chen 0002 Denny Wu Difan Zou Taiji Suzuki How Does Label Noise Gradient Descent Improve Generalization in the Low SNR Regime? 2025 NeurIPS http://papers.nips.cc/paper_files/paper/2025/hash/ffab50f3cad7cb5733ca324e5be20976-Abstract-Conference.html conf/nips/2025 db/conf/nips/neurips2025.html#HuangHSCWZS25 Naoki Nishikawa Rei Higuchi Taiji Suzuki Degrees of Freedom for Linear Attention: Distilling Softmax Attention with Optimal Feature Efficiency. 2025 NeurIPS http://papers.nips.cc/paper_files/paper/2025/hash/c98ef086dc70d528e1c1aa1e66893365-Abstract-Conference.html conf/nips/2025 db/conf/nips/neurips2025.html#NishikawaHS25 Shintaro Wakasugi Taiji Suzuki State Size Independent Statistical Error Bound for Discrete Diffusion Models. 2025 NeurIPS http://papers.nips.cc/paper_files/paper/2025/hash/caabce23491e5418f4c9ea348359fcd2-Abstract-Conference.html conf/nips/2025 db/conf/nips/neurips2025.html#WakasugiS25

Juno Kim Dimitri Meunier Arthur Gretton Taiji Suzuki Zhu Li Optimality and Adaptivity of Deep Neural Features for Instrumental Variable Regression. 2025 January abs/2501.04898 CoRR https://doi.org/10.48550/arXiv.2501.04898 db/journals/corr/corr2501.html#abs-2501-04898 streams/journals/corr

Juno Kim Denny Wu Jason Lee Taiji Suzuki Metastable Dynamics of Chain-of-Thought Reasoning: Provable Benefits of Search, RL and Distillation. 2025 February abs/2502.01694 CoRR https://doi.org/10.48550/arXiv.2502.01694 db/journals/corr/corr2502.html#abs-2502-01694 streams/journals/corr

Ryotaro Kawata Kazusato Oko Atsushi Nitanda Taiji Suzuki Direct Distributional Optimization for Provable Alignment of Diffusion Models. 2025 February abs/2502.02954 CoRR https://doi.org/10.48550/arXiv.2502.02954 db/journals/corr/corr2502.html#abs-2502-02954 streams/journals/corr

Atsushi Nitanda Anzelle Lee Damian Tan Xing Kai Mizuki Sakaguchi Taiji Suzuki Propagation of Chaos for Mean-Field Langevin Dynamics and its Application to Model Ensemble. 2025 February abs/2502.05784 CoRR https://doi.org/10.48550/arXiv.2502.05784 db/journals/corr/corr2502.html#abs-2502-05784 streams/journals/corr

Rei Higuchi Ryotaro Kawata Naoki Nishikawa Kazusato Oko Shoichiro Yamaguchi Sosuke Kobayashi Seiya Tokui Kohei Hayashi Daisuke Okanohara Taiji Suzuki When Does Metadata Conditioning (NOT) Work for Language Model Pre-Training? A Study with Context-Free Grammars. 2025 April abs/2504.17562 CoRR https://doi.org/10.48550/arXiv.2504.17562 db/journals/corr/corr2504.html#abs-2504-17562 streams/journals/corr

Rom N. Parnichkun Neehal Tumma Armin W. Thomas Alessandro Moro Qi An 0001 Taiji Suzuki Atsushi Yamashita Michael Poli Stefano Massaroli Quantifying Memory Utilization with Effective State-Size. 2025 April abs/2504.19561 CoRR https://doi.org/10.48550/arXiv.2504.19561 db/journals/corr/corr2504.html#abs-2504-19561 streams/journals/corr

Rei Higuchi Taiji Suzuki Direct Density Ratio Optimization: A Statistically Consistent Approach to Aligning Large Language Models. 2025 May abs/2505.07558 CoRR https://doi.org/10.48550/arXiv.2505.07558 db/journals/corr/corr2505.html#abs-2505-07558 streams/journals/corr

Andi Han Wei Huang 0034 Zhanpeng Zhou Gang Niu 0001 Wuyang Chen 0001 Junchi Yan Akiko Takeda Taiji Suzuki On the Role of Label Noise in the Feature Learning Process. 2025 May abs/2505.18909 CoRR https://doi.org/10.48550/arXiv.2505.18909 db/journals/corr/corr2505.html#abs-2505-18909 streams/journals/corr

Ryotaro Kawata Kohsei Matsutani Yuri Kinoshita Naoki Nishikawa Taiji Suzuki Mixture of Experts Provably Detect and Learn the Latent Cluster Structure in Gradient-Based Learning. 2025 June abs/2506.01656 CoRR https://doi.org/10.48550/arXiv.2506.01656 db/journals/corr/corr2506.html#abs-2506-01656 streams/journals/corr

Yilan Chen 0002 Zhichao Wang Wei Huang 0034 Andi Han Taiji Suzuki Arya Mazumdar Generalization Bound of Gradient Flow through Training Trajectory and Data-dependent Kernel. 2025 June abs/2506.11357 CoRR https://doi.org/10.48550/arXiv.2506.11357 db/journals/corr/corr2506.html#abs-2506-11357 streams/journals/corr

Naoki Nishikawa Rei Higuchi Taiji Suzuki Degrees of Freedom for Linear Attention: Distilling Softmax Attention with Optimal Feature Efficiency. 2025 July abs/2507.03340 CoRR https://doi.org/10.48550/arXiv.2507.03340 db/journals/corr/corr2507.html#abs-2507-03340 streams/journals/corr

Dake Bu Wei Huang 0034 Andi Han Atsushi Nitanda Qingfu Zhang 0001 Hau-San Wong Taiji Suzuki Provable In-Context Vector Arithmetic via Retrieving Task Concepts. 2025 August abs/2508.09820 CoRR https://doi.org/10.48550/arXiv.2508.09820 db/journals/corr/corr2508.html#abs-2508-09820 streams/journals/corr

Jiarui Jiang Wei Huang 0034 Miao Zhang 0022 Taiji Suzuki Liqiang Nie Trained Mamba Emulates Online Gradient Descent in In-Context Linear Regression. 2025 September abs/2509.23779 CoRR https://doi.org/10.48550/arXiv.2509.23779 db/journals/corr/corr2509.html#abs-2509-23779 streams/journals/corr

Kento Kuwataka Taiji Suzuki Test time training enhances in-context learning of nonlinear functions. 2025 September abs/2509.25741 CoRR https://doi.org/10.48550/arXiv.2509.25741 db/journals/corr/corr2509.html#abs-2509-25741 streams/journals/corr

Tomoya Wakayama Taiji Suzuki In-Context Learning Is Provably Bayesian Inference: A Generalization Theory for Meta-Learning. 2025 October abs/2510.10981 CoRR https://doi.org/10.48550/arXiv.2510.10981 db/journals/corr/corr2510.html#abs-2510-10981 streams/journals/corr

Junsoo Oh Wei Huang 0034 Taiji Suzuki Mamba Can Learn Low-Dimensional Targets In-Context via Test-Time Feature Learning. 2025 October abs/2510.12026 CoRR https://doi.org/10.48550/arXiv.2510.12026 db/journals/corr/corr2510.html#abs-2510-12026 streams/journals/corr

Wei Huang 0034 Andi Han Yujin Song Yilan Chen 0002 Denny Wu Difan Zou Taiji Suzuki How Does Label Noise Gradient Descent Improve Generalization in the Low SNR Regime? 2025 October abs/2510.17526 CoRR https://doi.org/10.48550/arXiv.2510.17526 db/journals/corr/corr2510.html#abs-2510-17526 streams/journals/corr

Dake Bu Wei Huang 0034 Andi Han Atsushi Nitanda Bo Xue 0004 Qingfu Zhang 0001 Hau-San Wong Taiji Suzuki Consistency Is Not Always Correct: Towards Understanding the Role of Exploration in Post-Training Reasoning. 2025 November abs/2511.07368 CoRR https://doi.org/10.48550/arXiv.2511.07368 db/journals/corr/corr2511.html#abs-2511-07368 streams/journals/corr

Dake Bu Wei Huang 0034 Andi Han Atsushi Nitanda Hau-San Wong Qingfu Zhang 0001 Taiji Suzuki Provable Benefit of Curriculum in Transformer Tree-Reasoning Post-Training. 2025 November abs/2511.07372 CoRR https://doi.org/10.48550/arXiv.2511.07372 db/journals/corr/corr2511.html#abs-2511-07372 streams/journals/corr

Zonghao Chen Atsushi Nitanda Arthur Gretton Taiji Suzuki Towards a Unified Analysis of Neural Networks in Nonparametric Instrumental Variable Regression: Optimization and Generalization. 2025 November abs/2511.14710 CoRR https://doi.org/10.48550/arXiv.2511.14710 db/journals/corr/corr2511.html#abs-2511-14710 streams/journals/corr

Dragos Secrieru Garyk Brixi Yoshua Bengio Taiji Suzuki Michael Poli Stefano Massaroli Sliding Window Recurrences for Sequence Models. 2025 December abs/2512.13921 CoRR https://doi.org/10.48550/arXiv.2512.13921 db/journals/corr/corr2512.html#abs-2512-13921 streams/journals/corr

Ryotaro Kawata Yujin Song Alberto Bietti Naoki Nishikawa Taiji Suzuki Samuel Vaiter Denny Wu From Shortcut to Induction Head: How Data Diversity Shapes Algorithm Selection in Transformers. 2025 December abs/2512.18634 CoRR https://doi.org/10.48550/arXiv.2512.18634 db/journals/corr/corr2512.html#abs-2512-18634 streams/journals/corr

Kazusato Oko Yujin Song Taiji Suzuki Denny Wu Learning sum of diverse features: computational hardness and efficient gradient-based training for ridge combinations. 4009-4081 2024 COLT https://proceedings.mlr.press/v247/oko24a.html conf/colt/2024 db/conf/colt/colt2024.html#OkoSSW24 Yuka Hashimoto Sho Sonoda Isao Ishikawa Atsushi Nitanda Taiji Suzuki Koopman-based generalization bound: New aspect for full-rank weights. 2024 ICLR https://openreview.net/forum?id=JN7TcCm9LF conf/iclr/2024 db/conf/iclr/iclr2024.html#HashimotoSINS24 Wei Huang 0034 Ye Shi 0001 Zhongyi Cai Taiji Suzuki Understanding Convergence and Generalization in Federated Learning through Feature Learning Theory. 2024 ICLR https://openreview.net/forum?id=EcetCr4trp conf/iclr/2024 db/conf/iclr/iclr2024.html#Huang0CS24 Juno Kim Kakei Yamamoto Kazusato Oko Zhuoran Yang Taiji Suzuki Symmetric Mean-field Langevin Dynamics for Distributional Minimax Problems. 2024 ICLR https://openreview.net/forum?id=YItWKZci78 conf/iclr/2024 db/conf/iclr/iclr2024.html#KimYOYS24 Yuto Nishimura Taiji Suzuki Minimax optimality of convolutional neural networks for infinite dimensional input-output problems and separation from kernel methods. 2024 ICLR https://openreview.net/forum?id=EW8ZExRZkJ conf/iclr/2024 db/conf/iclr/iclr2024.html#NishimuraS24 Atsushi Nitanda Kazusato Oko Taiji Suzuki Denny Wu Improved statistical and computational complexity of the mean-field Langevin dynamics under structured data. 2024 ICLR https://openreview.net/forum?id=Of2nEDc4s7 conf/iclr/2024 db/conf/iclr/iclr2024.html#NitandaOSW24 Keita Suzuki Taiji Suzuki Optimal criterion for feature learning of two-layer linear neural network in high dimensional interpolation regime. 2024 ICLR https://openreview.net/forum?id=Jc0FssXh2R conf/iclr/2024 db/conf/iclr/iclr2024.html#SuzukiS24 Dake Bu Wei Huang 0034 Taiji Suzuki Ji Cheng 0001 Qingfu Zhang 0001 Zhiqiang Xu 0003 Hau-San Wong Provably Neural Active Learning Succeeds via Prioritizing Perplexing Samples. 2024 ICML https://proceedings.mlr.press/v235/bu24a.html https://openreview.net/forum?id=kzz0kn546b conf/icml/2024 db/conf/icml/icml2024.html#BuHSC0XW24 4642-4695 Yihang Chen 0003 Fanghui Liu 0001 Taiji Suzuki Volkan Cevher High-Dimensional Kernel Methods under Covariate Shift: Data-Dependent Implicit Regularization. 2024 ICML https://proceedings.mlr.press/v235/chen24aa.html https://openreview.net/forum?id=bBzlapzeR1 conf/icml/2024 db/conf/icml/icml2024.html#Chen0SC24 7081-7102 Juno Kim Taiji Suzuki Transformers Learn Nonlinear Features In Context: Nonconvex Mean-field Dynamics on the Attention Landscape. 2024 ICML https://proceedings.mlr.press/v235/kim24af.html https://openreview.net/forum?id=xm2lU7tteQ conf/icml/2024 db/conf/icml/icml2024.html#KimS24 24527-24561 Kazusato Oko Shunta Akiyama Denny Wu Tomoya Murata Taiji Suzuki SILVER: Single-loop variance reduction and application to federated learning. 2024 ICML https://proceedings.mlr.press/v235/oko24a.html https://openreview.net/forum?id=pOgMluzEIH conf/icml/2024 db/conf/icml/icml2024.html#OkoAWMS24 38683-38739 Rom N. Parnichkun Stefano Massaroli Alessandro Moro Jimmy T. H. Smith Ramin M. Hasani Mathias Lechner Qi An 0001 Christopher Ré Hajime Asama Stefano Ermon Taiji Suzuki Michael Poli Atsushi Yamashita State-Free Inference of State-Space Models: The *Transfer Function* Approach. 2024 ICML https://proceedings.mlr.press/v235/parnichkun24a.html https://openreview.net/forum?id=DwwI9L67B5 conf/icml/2024 db/conf/icml/icml2024.html#ParnichkunMMSHL24 39834-39860 Michael Poli Armin W. Thomas Eric Nguyen Pragaash Ponnusamy Björn Deiseroth Kristian Kersting Taiji Suzuki Brian L. Hie Stefano Ermon Christopher Ré Ce Zhang 0001 Stefano Massaroli Mechanistic Design and Scaling of Hybrid Architectures. 2024 ICML https://proceedings.mlr.press/v235/poli24a.html https://openreview.net/forum?id=GDp7Gyd9nf conf/icml/2024 db/conf/icml/icml2024.html#PoliTNPDKSHER0M24 40908-40950 Michael Eli Sander Raja Giryes Taiji Suzuki Mathieu Blondel Gabriel Peyré How do Transformers Perform In-Context Autoregressive Learning ? 2024 ICML https://proceedings.mlr.press/v235/sander24a.html https://openreview.net/forum?id=kZbTkpnafR conf/icml/2024 db/conf/icml/icml2024.html#SanderGSBP24 43235-43254 Shokichi Takakura Taiji Suzuki Mean-field Analysis on Two-layer Neural Networks from a Kernel Perspective. 2024 ICML https://proceedings.mlr.press/v235/takakura24a.html https://openreview.net/forum?id=mphq2jMFLZ conf/icml/2024 db/conf/icml/icml2024.html#TakakuraS24 47475-47509 Kakei Yamamoto Kazusato Oko Zhuoran Yang Taiji Suzuki Mean Field Langevin Actor-Critic: Faster Convergence and Global Optimality beyond Lazy Learning. 2024 ICML https://proceedings.mlr.press/v235/yamamoto24a.html https://openreview.net/forum?id=FOJE1kRcHG conf/icml/2024 db/conf/icml/icml2024.html#YamamotoOYS24 55706-55738 Kishan Wimalawarne Taro Sawaki Motokiyo Hirayama Takanobu Kawahara Taiji Suzuki Graph Polynomial Convolution Models for Node Classification of Non-Homophilous Graphs. 856-863 2024 ICMLA https://doi.org/10.1109/ICMLA61862.2024.00124 conf/icmla/2024 db/conf/icmla/icmla2024.html#WimalawarneSHKS24 streams/conf/icmla Dake Bu Wei Huang 0034 Andi Han Atsushi Nitanda Taiji Suzuki Qingfu Zhang 0001 Hau-San Wong Provably Transformers Harness Multi-Concept Word Semantics for Efficient In-Context Learning. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/73efab19ebde03ff0958f4f155483f57-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#BuHHNS0W24 Wei Huang 0034 Andi Han Yongqiang Chen 0002 Yuan Cao 0006 Zhiqiang Xu 0003 Taiji Suzuki On the Comparison between Multi-modal and Single-modal Contrastive Learning. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/94840c41497ead6a84f493f029eba7fa-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#HuangHCCXS24 Jiarui Jiang Wei Huang 0034 Miao Zhang 0022 Taiji Suzuki Liqiang Nie Unveil Benign Overfitting for Transformer in Vision: Training Dynamics, Convergence, and Generalization. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/f49287371916715b9209fa41a275851e-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#Jiang0ZSN24 Juno Kim Tai Nakamaki Taiji Suzuki Transformers are Minimax Optimal Nonparametric In-Context Learners. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/c11daad0a48ea5f3c5c6390c7b060720-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#KimNS24 Jason D. Lee Kazusato Oko Taiji Suzuki Denny Wu Neural network learns low-dimensional polynomials with SGD near the information-theoretic limit. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/6bd5fca2074dcd9ede9de50f71f7ec28-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#LeeOSW24 Kazusato Oko Yujin Song Taiji Suzuki Denny Wu Pretrained Transformer Efficiently Learns Low-Dimensional Target Functions In-Context. 2024 NeurIPS http://papers.nips.cc/paper_files/paper/2024/hash/8d6c356c82252032a20c749dda36f2d8-Abstract-Conference.html conf/nips/2024 db/conf/nips/neurips2024.html#OkoSSW24 Kazuki Uematsu Kosuke Haruki Taiji Suzuki Mitsuhiro Kimura Takahiro Takimoto Hideyuki Nakagawa Dimensionality-Induced Information Loss of Outliers in Deep Neural Networks. 144-160 2024 ECML/PKDD (1) https://doi.org/10.1007/978-3-031-70341-6_9 conf/pkdd/2024-1 db/conf/pkdd/pkdd2024-1.html#UematsuHSKTN24 streams/conf/pkdd

Juno Kim Taiji Suzuki Transformers Learn Nonlinear Features In Context: Nonconvex Mean-field Dynamics on the Attention Landscape. 2024 abs/2402.01258 CoRR https://doi.org/10.48550/arXiv.2402.01258 db/journals/corr/corr2402.html#abs-2402-01258

Michael E. Sander Raja Giryes Taiji Suzuki Mathieu Blondel Gabriel Peyré How do Transformers perform In-Context Autoregressive Learning? 2024 abs/2402.05787 CoRR https://doi.org/10.48550/arXiv.2402.05787 db/journals/corr/corr2402.html#abs-2402-05787

Shokichi Takakura Taiji Suzuki Mean-field Analysis on Two-layer Neural Networks from a Kernel Perspective. 2024 abs/2403.14917 CoRR https://doi.org/10.48550/arXiv.2403.14917 db/journals/corr/corr2403.html#abs-2403-14917

Michael Poli Armin W. Thomas Eric Nguyen Pragaash Ponnusamy Björn Deiseroth Kristian Kersting Taiji Suzuki Brian L. Hie Stefano Ermon Christopher Ré Ce Zhang 0001 Stefano Massaroli Mechanistic Design and Scaling of Hybrid Architectures. 2024 abs/2403.17844 CoRR https://doi.org/10.48550/arXiv.2403.17844 db/journals/corr/corr2403.html#abs-2403-17844

Toshimitsu Uesaka Taiji Suzuki Yuhta Takida Chieh-Hsin Lai Naoki Murata Yuki Mitsufuji Understanding Multimodal Contrastive Learning Through Pointwise Mutual Information. 2024 abs/2404.19228 CoRR https://doi.org/10.48550/arXiv.2404.19228 db/journals/corr/corr2404.html#abs-2404-19228

Rom N. Parnichkun Stefano Massaroli Alessandro Moro Jimmy T. H. Smith Ramin M. Hasani Mathias Lechner Qi An 0001 Christopher Ré Hajime Asama Stefano Ermon Taiji Suzuki Atsushi Yamashita Michael Poli State-Free Inference of State-Space Models: The Transfer Function Approach. 2024 abs/2405.06147 CoRR https://doi.org/10.48550/arXiv.2405.06147 db/journals/corr/corr2405.html#abs-2405-06147

Naoki Nishikawa Taiji Suzuki State Space Models are Comparable to Transformers in Estimating Functions with Dynamic Smoothness. 2024 abs/2405.19036 CoRR https://doi.org/10.48550/arXiv.2405.19036 db/journals/corr/corr2405.html#abs-2405-19036

Kenji Fukumizu Taiji Suzuki Noboru Isobe Kazusato Oko Masanori Koyama Flow matching achieves minimax optimal convergence. 2024 abs/2405.20879 CoRR https://doi.org/10.48550/arXiv.2405.20879 db/journals/corr/corr2405.html#abs-2405-20879

Jason D. Lee Kazusato Oko Taiji Suzuki Denny Wu Neural network learns low-dimensional polynomials with SGD near the information-theoretic limit. 2024 abs/2406.01581 CoRR https://doi.org/10.48550/arXiv.2406.01581 db/journals/corr/corr2406.html#abs-2406-01581

Yihang Chen 0003 Fanghui Liu 0001 Taiji Suzuki Volkan Cevher High-Dimensional Kernel Methods under Covariate Shift: Data-Dependent Implicit Regularization. 2024 abs/2406.03171 CoRR https://doi.org/10.48550/arXiv.2406.03171 db/journals/corr/corr2406.html#abs-2406-03171 streams/journals/corr

Dake Bu Wei Huang 0034 Taiji Suzuki Ji Cheng 0001 Qingfu Zhang 0001 Zhiqiang Xu 0003 Hau-San Wong Provably Neural Active Learning Succeeds via Prioritizing Perplexing Samples. 2024 abs/2406.03944 CoRR https://doi.org/10.48550/arXiv.2406.03944 db/journals/corr/corr2406.html#abs-2406-03944

Kazusato Oko Yujin Song Taiji Suzuki Denny Wu Learning sum of diverse features: computational hardness and efficient gradient-based training for ridge combinations. 2024 abs/2406.11828 CoRR https://doi.org/10.48550/arXiv.2406.11828 db/journals/corr/corr2406.html#abs-2406-11828 streams/journals/corr

Juno Kim Tai Nakamaki Taiji Suzuki Transformers are Minimax Optimal Nonparametric In-Context Learners. 2024 abs/2408.12186 CoRR https://doi.org/10.48550/arXiv.2408.12186 db/journals/corr/corr2408.html#abs-2408-12186 streams/journals/corr

Jiarui Jiang Wei Huang 0034 Miao Zhang 0022 Taiji Suzuki Liqiang Nie Unveil Benign Overfitting for Transformer in Vision: Training Dynamics, Convergence, and Generalization. 2024 abs/2409.19345 CoRR https://doi.org/10.48550/arXiv.2409.19345 db/journals/corr/corr2409.html#abs-2409-19345 streams/journals/corr

Bingrui Li Wei Huang 0034 Andi Han Zhanpeng Zhou Taiji Suzuki Jun Zhu 0001 Jianfei Chen 0001 On the Optimization and Generalization of Two-layer Transformers with Sign Gradient Descent. 2024 abs/2410.04870 CoRR https://doi.org/10.48550/arXiv.2410.04870 db/journals/corr/corr2410.html#abs-2410-04870 streams/journals/corr

Juno Kim Taiji Suzuki Transformers Provably Solve Parity Efficiently with Chain of Thought. 2024 abs/2410.08633 CoRR https://doi.org/10.48550/arXiv.2410.08633 db/journals/corr/corr2410.html#abs-2410-08633 streams/journals/corr

Kazuki Uematsu Kosuke Haruki Taiji Suzuki Mitsuhiro Kimura Takahiro Takimoto Hideyuki Nakagawa Dimensionality-induced information loss of outliers in deep neural networks. 2024 abs/2410.21656 CoRR https://doi.org/10.48550/arXiv.2410.21656 db/journals/corr/corr2410.html#abs-2410-21656 streams/journals/corr

Dake Bu Wei Huang 0034 Andi Han Atsushi Nitanda Taiji Suzuki Qingfu Zhang 0001 Hau-San Wong Provably Transformers Harness Multi-Concept Word Semantics for Efficient In-Context Learning. 2024 abs/2411.02199 CoRR https://doi.org/10.48550/arXiv.2411.02199 db/journals/corr/corr2411.html#abs-2411-02199 streams/journals/corr

Kazusato Oko Yujin Song Taiji Suzuki Denny Wu Pretrained transformer efficiently learns low-dimensional target functions in-context. 2024 abs/2411.02544 CoRR https://doi.org/10.48550/arXiv.2411.02544 db/journals/corr/corr2411.html#abs-2411-02544 streams/journals/corr

Wei Huang 0034 Andi Han Yongqiang Chen 0002 Yuan Cao 0006 Zhiqiang Xu 0003 Taiji Suzuki On the Comparison between Multi-modal and Single-modal Contrastive Learning. 2024 abs/2411.02837 CoRR https://doi.org/10.48550/arXiv.2411.02837 db/journals/corr/corr2411.html#abs-2411-02837 streams/journals/corr

Shunta Akiyama Taiji Suzuki Excess Risk of Two-Layer ReLU Neural Networks in Teacher-Student Settings and its Superiority to Kernel Methods. 2023 ICLR https://openreview.net/forum?id=6doXHqwMayf conf/iclr/2023 db/conf/iclr/iclr2023.html#AkiyamaS23 Taiji Suzuki Atsushi Nitanda Denny Wu Uniform-in-time propagation of chaos for the mean-field gradient Langevin dynamics. 2023 ICLR https://openreview.net/forum?id=_JScUk9TBUn conf/iclr/2023 db/conf/iclr/iclr2023.html#SuzukiNW23 Tomoya Murata Taiji Suzuki DIFF2: Differential Private Optimization via Gradient Differences for Nonconvex Distributed Learning. 25523-25548 2023 ICML https://proceedings.mlr.press/v202/murata23b.html conf/icml/2023 db/conf/icml/icml2023.html#MurataS23 Atsushi Nitanda Kazusato Oko Denny Wu Nobuhito Takenouchi Taiji Suzuki Primal and Dual Analysis of Entropic Fictitious Play for Finite-sum Problems. 26266-26282 2023 ICML https://proceedings.mlr.press/v202/nitanda23a.html conf/icml/2023 db/conf/icml/icml2023.html#NitandaOWTS23 Kazusato Oko Shunta Akiyama Taiji Suzuki Diffusion Models are Minimax Optimal Distribution Estimators. 26517-26582 2023 ICML https://proceedings.mlr.press/v202/oko23a.html conf/icml/2023 db/conf/icml/icml2023.html#OkoAS23 Atsushi Suzuki 0002 Atsushi Nitanda Taiji Suzuki Jing Wang 0023 Feng Tian 0006 Kenji Yamanishi Tight and fast generalization error bound of graph embedding in metric space. 33268-33284 2023 ICML https://proceedings.mlr.press/v202/suzuki23a.html conf/icml/2023 db/conf/icml/icml2023.html#SuzukiNS00Y23 Shokichi Takakura Taiji Suzuki Approximation and Estimation Ability of Transformers for Sequence-to-Sequence Functions with Infinite Dimensional Input. 33416-33447 2023 ICML https://proceedings.mlr.press/v202/takakura23a.html conf/icml/2023 db/conf/icml/icml2023.html#TakakuraS23 Shuhei Nitta Taiji Suzuki Albert Rodríguez Mulet Atsushi Yaguchi Ryusuke Hirai Scalable Federated Learning for Clients with Different Input Image Sizes and Numbers of Output Categories. 764-769 2023 ICMLA https://doi.org/10.1109/ICMLA58977.2023.00111 conf/icmla/2023 db/conf/icmla/icmla2023.html#NittaSMYH23 Hiroaki Kingetsu Kenichi Kobayashi 0001 Taiji Suzuki Neural Network Module Decomposition and Recomposition with Superimposed Masks. 1-10 2023 IJCNN https://doi.org/10.1109/IJCNN54540.2023.10191417 conf/ijcnn/2023 db/conf/ijcnn/ijcnn2023.html#KingetsuKS23 Jimmy Ba Murat A. Erdogdu Taiji Suzuki Zhichao Wang Denny Wu Learning in the Presence of Low-dimensional Structure: A Spiked Random Matrix Perspective. 2023 NeurIPS http://papers.nips.cc/paper_files/paper/2023/hash/38a1671ab0747b6ffe4d1c6ef117a3a9-Abstract-Conference.html conf/nips/2023 db/conf/nips/neurips2023.html#BaESWW23 Alireza Mousavi-Hosseini Denny Wu Taiji Suzuki Murat A. Erdogdu Gradient-Based Feature Learning under Structured Data. 2023 NeurIPS http://papers.nips.cc/paper_files/paper/2023/hash/e21955c93dede886af1d0d362c756757-Abstract-Conference.html conf/nips/2023 db/conf/nips/neurips2023.html#HosseiniWSE23 Taiji Suzuki Denny Wu Atsushi Nitanda Mean-field Langevin dynamics: Time-space discretization, stochastic gradient, and variance reduction. 2023 NeurIPS http://papers.nips.cc/paper_files/paper/2023/hash/32133a6a24d6554263d3584e3ac10faa-Abstract-Conference.html conf/nips/2023 db/conf/nips/neurips2023.html#SuzukiWN23 Taiji Suzuki Denny Wu Kazusato Oko Atsushi Nitanda Feature learning via mean-field Langevin dynamics: classifying sparse parities and beyond. 2023 NeurIPS http://papers.nips.cc/paper_files/paper/2023/hash/6cc321baf0a8611b1d1bdbd18822667b-Abstract-Conference.html conf/nips/2023 db/conf/nips/neurips2023.html#SuzukiWON23

Tomoya Murata Taiji Suzuki DIFF2: Differential Private Optimization via Gradient Differences for Nonconvex Distributed Learning. 2023 abs/2302.03884 CoRR https://doi.org/10.48550/arXiv.2302.03884 db/journals/corr/corr2302.html#abs-2302-03884

Yuka Hashimoto Sho Sonoda Isao Ishikawa Atsushi Nitanda Taiji Suzuki Koopman-Based Bound for Generalization: New Aspect of Neural Networks Regarding Nonlinear Noise Filtering. 2023 abs/2302.05825 CoRR https://doi.org/10.48550/arXiv.2302.05825 db/journals/corr/corr2302.html#abs-2302-05825

Kazusato Oko Shunta Akiyama Taiji Suzuki Diffusion Models are Minimax Optimal Distribution Estimators. 2023 abs/2303.01861 CoRR https://doi.org/10.48550/arXiv.2303.01861 db/journals/corr/corr2303.html#abs-2303-01861

Atsushi Nitanda Kazusato Oko Denny Wu Nobuhito Takenouchi Taiji Suzuki Primal and Dual Analysis of Entropic Fictitious Play for Finite-sum Problems. 2023 abs/2303.02957 CoRR https://doi.org/10.48550/arXiv.2303.02957 db/journals/corr/corr2303.html#abs-2303-02957

Atsushi Suzuki 0002 Atsushi Nitanda Taiji Suzuki Jing Wang 0023 Feng Tian 0006 Kenji Yamanishi Tight and fast generalization error bound of graph embedding in metric space. 2023 abs/2305.07971 CoRR https://doi.org/10.48550/arXiv.2305.07971 db/journals/corr/corr2305.html#abs-2305-07971

Shokichi Takakura Taiji Suzuki Approximation and Estimation Ability of Transformers for Sequence-to-Sequence Functions with Infinite Dimensional Input. 2023 abs/2305.18699 CoRR https://doi.org/10.48550/arXiv.2305.18699 db/journals/corr/corr2305.html#abs-2305-18699

Taiji Suzuki Denny Wu Atsushi Nitanda Convergence of mean-field Langevin dynamics: Time and space discretization, stochastic gradient, and variance reduction. 2023 abs/2306.07221 CoRR https://doi.org/10.48550/arXiv.2306.07221 db/journals/corr/corr2306.html#abs-2306-07221

Wei Huang 0034 Yuan Cao 0006 Haonan Wang Xin Cao 0001 Taiji Suzuki Graph Neural Networks Provably Benefit from Structural Information: A Feature Learning Perspective. 2023 abs/2306.13926 CoRR https://doi.org/10.48550/arXiv.2306.13926 db/journals/corr/corr2306.html#abs-2306-13926

Kishan Wimalawarne Taiji Suzuki Sophie Langer Learning Green's Function Efficiently Using Low-Rank Approximations. 2023 abs/2308.00350 CoRR https://doi.org/10.48550/arXiv.2308.00350 db/journals/corr/corr2308.html#abs-2308-00350

Alireza Mousavi-Hosseini Denny Wu Taiji Suzuki Murat A. Erdogdu Gradient-Based Feature Learning under Structured Data. 2023 abs/2309.03843 CoRR https://doi.org/10.48550/arXiv.2309.03843 db/journals/corr/corr2309.html#abs-2309-03843

Shuhei Nitta Taiji Suzuki Albert Rodríguez Mulet Atsushi Yaguchi Ryusuke Hirai Scalable Federated Learning for Clients with Different Input Image Sizes and Numbers of Output Categories. 2023 abs/2311.08716 CoRR https://doi.org/10.48550/arXiv.2311.08716 db/journals/corr/corr2311.html#abs-2311-08716

Chihiro Watanabe Taiji Suzuki Deep two-way matrix reordering for relational data analysis. 303-315 2022 146 Neural Networks https://doi.org/10.1016/j.neunet.2021.11.028 db/journals/nn/nn146.html#WatanabeS22

Kishan Wimalawarne Taiji Suzuki Layer-wise Adaptive Graph Convolution Networks Using Generalized Pagerank. 1117-1132 2022 ACML https://proceedings.mlr.press/v189/wimalawarne23a.html conf/acml/2022 db/conf/acml/acml2022.html#WimalawarneS22 Atsushi Nitanda Denny Wu Taiji Suzuki Convex Analysis of the Mean Field Langevin Dynamics. 9741-9757 2022 AISTATS https://proceedings.mlr.press/v151/nitanda22a.html conf/aistats/2022 db/conf/aistats/aistats2022.html#NitandaWS22 Boris Muzellec Kanji Sato Mathurin Massias Taiji Suzuki Dimension-free convergence rates for gradient Langevin dynamics in RKHS. 1356-1420 2022 COLT https://proceedings.mlr.press/v178/muzellec22a.html conf/colt/2022 db/conf/colt/colt2022.html#MuzellecSMS22 Jimmy Ba Murat A. Erdogdu Marzyeh Ghassemi Shengyang Sun Taiji Suzuki Denny Wu Tianzong Zhang Understanding the Variance Collapse of SVGD in High Dimensions. 2022 ICLR https://openreview.net/forum?id=Qycd9j5Qp9J conf/iclr/2022 db/conf/iclr/iclr2022.html#BaEGSSWZ22 Kazusato Oko Taiji Suzuki Atsushi Nitanda Denny Wu Particle Stochastic Dual Coordinate Ascent: Exponential convergent algorithm for mean field neural network optimization. 2022 ICLR https://openreview.net/forum?id=PQQp7AJwz3 conf/iclr/2022 db/conf/iclr/iclr2022.html#OkoSNW22 Sho Okumoto Taiji Suzuki Learnability of convolutional neural networks for infinite dimensional input via mixed and anisotropic smoothness. 2022 ICLR https://openreview.net/forum?id=dgxFTxuJ50e conf/iclr/2022 db/conf/iclr/iclr2022.html#OkumotoS22 Chenyuan Xu Kosuke Haruki Taiji Suzuki Masahiro Ozawa Kazuki Uematsu Ryuji Sakai Data-Parallel Momentum Diagonal Empirical Fisher (DP-MDEF):Adaptive Gradient Method is Affected by Hessian Approximation and Multi-Class Data. 1397-1404 2022 ICMLA https://doi.org/10.1109/ICMLA55696.2022.00221 conf/icmla/2022 db/conf/icmla/icmla2022.html#XuHSOUS22 Kengo Machida Kuniaki Uto Koichi Shinoda Taiji Suzuki MSR-DARTS: Minimum Stable Rank of Differentiable Architecture Search. 1-9 2022 IJCNN https://doi.org/10.1109/IJCNN55064.2022.9892751 conf/ijcnn/2022 db/conf/ijcnn/ijcnn2022.html#MachidaUSS22 Jimmy Ba Murat A. Erdogdu Taiji Suzuki Zhichao Wang Denny Wu Greg Yang High-dimensional Asymptotics of Feature Learning: How One Gradient Step Improves the Representation. 2022 conf/nips/2022 NeurIPS http://papers.nips.cc/paper_files/paper/2022/hash/f7e7fabd73b3df96c54a320862afcb78-Abstract-Conference.html db/conf/nips/neurips2022.html#BaESWWY22 Yuri Kinoshita Taiji Suzuki Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization. 2022 conf/nips/2022 NeurIPS http://papers.nips.cc/paper_files/paper/2022/hash/78e839f96568985d18463044a064ea0f-Abstract-Conference.html db/conf/nips/neurips2022.html#KinoshitaS22 Tomoya Murata Taiji Suzuki Escaping Saddle Points with Bias-Variance Reduced Local Perturbed SGD for Communication Efficient Nonconvex Distributed Learning. 2022 conf/nips/2022 NeurIPS http://papers.nips.cc/paper_files/paper/2022/hash/20bd42d82998bc61732c00452228e814-Abstract-Conference.html db/conf/nips/neurips2022.html#MurataS22 Naoki Nishikawa Taiji Suzuki Atsushi Nitanda Denny Wu Two-layer neural network on infinite dimensional data: global optimization guarantee in the mean-field regime. 2022 conf/nips/2022 NeurIPS http://papers.nips.cc/paper_files/paper/2022/hash/d2155b1f7eb42350d7bc3013eefe5480-Abstract-Conference.html db/conf/nips/neurips2022.html#NishikawaSNW22 Hiroaki Mikami Kenji Fukumizu Shogo Murai Shuji Suzuki Yuta Kikuchi Taiji Suzuki Shin-ichi Maeda Kohei Hayashi A Scaling Law for Syn2real Transfer: How Much Is Your Pre-training Effective? 477-492 2022 ECML/PKDD (3) https://doi.org/10.1007/978-3-031-26409-2_29 conf/pkdd/2022-3 db/conf/pkdd/pkdd2022-3.html#MikamiFMSKSMH22

Atsushi Nitanda Denny Wu Taiji Suzuki Convex Analysis of the Mean Field Langevin Dynamics. 2022 abs/2201.10469 CoRR https://arxiv.org/abs/2201.10469 db/journals/corr/corr2201.html#abs-2201-10469

Tomoya Murata Taiji Suzuki Escaping Saddle Points with Bias-Variance Reduced Local Perturbed SGD for Communication Efficient Nonconvex Distributed Learning. 2022 abs/2202.06083 CoRR https://arxiv.org/abs/2202.06083 db/journals/corr/corr2202.html#abs-2202-06083

Yuri Kinoshita Taiji Suzuki Improved Convergence Rate of Stochastic Gradient Langevin Dynamics with Variance Reduction and its Application to Optimization. 2022 abs/2203.16217 CoRR https://doi.org/10.48550/arXiv.2203.16217 db/journals/corr/corr2203.html#abs-2203-16217

Jimmy Ba Murat A. Erdogdu Taiji Suzuki Zhichao Wang Denny Wu Greg Yang High-dimensional Asymptotics of Feature Learning: How One Gradient Step Improves the Representation. 2022 abs/2205.01445 CoRR https://doi.org/10.48550/arXiv.2205.01445 db/journals/corr/corr2205.html#abs-2205-01445

Shunta Akiyama Taiji Suzuki Excess Risk of Two-Layer ReLU Neural Networks in Teacher-Student Settings and its Superiority to Kernel Methods. 2022 abs/2205.14818 CoRR https://doi.org/10.48550/arXiv.2205.14818 db/journals/corr/corr2205.html#abs-2205-14818

Kazusato Oko Shunta Akiyama Tomoya Murata Taiji Suzuki Versatile Single-Loop Method for Gradient Estimator: First and Second Order Optimality, and its Application to Federated Learning. 2022 abs/2209.00361 CoRR https://doi.org/10.48550/arXiv.2209.00361 db/journals/corr/corr2209.html#abs-2209-00361

Kishan Wimalawarne Taiji Suzuki Graph Polynomial Convolution Models for Node Classification of Non-Homophilous Graphs. 2022 abs/2209.05020 CoRR https://doi.org/10.48550/arXiv.2209.05020 db/journals/corr/corr2209.html#abs-2209-05020

Chihiro Watanabe Taiji Suzuki Goodness-of-fit test for latent block models. 107090 2021 154 Comput. Stat. Data Anal. https://doi.org/10.1016/j.csda.2020.107090 db/journals/csda/csda154.html#WatanabeS21

Atsushi Nitanda Tomoya Murata Taiji Suzuki Sharp characterization of optimal minibatch size for stochastic finite sum convex optimization. 2513-2539 2021 63 Knowl. Inf. Syst. 9 https://doi.org/10.1007/s10115-021-01593-1 db/journals/kais/kais63.html#NitandaMS21

Shingo Yashima Atsushi Nitanda Taiji Suzuki Exponential Convergence Rates of Classification Errors on Learning with SGD and Random Features. 1954-1962 2021 AISTATS http://proceedings.mlr.press/v130/yashima21a.html conf/aistats/2021 db/conf/aistats/aistats2021.html#YashimaNS21 Tomoya Murata Taiji Suzuki Gradient Descent in RKHS with Importance Labeling. 1981-1989 2021 AISTATS http://proceedings.mlr.press/v130/murata21a.html conf/aistats/2021 db/conf/aistats/aistats2021.html#MurataS21 Shun-ichi Amari Jimmy Ba Roger Baker Grosse Xuechen Li 0005 Atsushi Nitanda Taiji Suzuki Denny Wu Ji Xu When does preconditioning help or hurt generalization? 2021 ICLR https://openreview.net/forum?id=S724o4_WB3 conf/iclr/2021 db/conf/iclr/iclr2021.html#AmariBGLNSWX21 Atsushi Nitanda Taiji Suzuki Optimal Rates for Averaged Stochastic Gradient Descent under Neural Tangent Kernel Regime. 2021 ICLR https://openreview.net/forum?id=PULSD5qI2N1 conf/iclr/2021 db/conf/iclr/iclr2021.html#NitandaS21 Taiji Suzuki Shunta Akiyama Benefit of deep learning with non-convex noisy gradient descent: Provable excess risk bound and superiority to kernel methods. 2021 ICLR https://openreview.net/forum?id=2m0g1wEafh conf/iclr/2021 db/conf/iclr/iclr2021.html#SuzukiA21 Shunta Akiyama Taiji Suzuki On Learnability via Gradient Method for Two-Layer ReLU Neural Networks in Teacher-Student Setting. 152-162 2021 ICML http://proceedings.mlr.press/v139/akiyama21a.html conf/icml/2021 db/conf/icml/icml2021.html#AkiyamaS21 Tomoya Murata Taiji Suzuki Bias-Variance Reduced Local SGD for Less Heterogeneous Federated Learning. 7872-7881 2021 ICML http://proceedings.mlr.press/v139/murata21a.html conf/icml/2021 db/conf/icml/icml2021.html#MurataS21 Akira Nakagawa Keizo Kato Taiji Suzuki Quantitative Understanding of VAE as a Non-linearly Scaled Isometric Embedding. 7916-7926 2021 ICML http://proceedings.mlr.press/v139/nakagawa21a.html conf/icml/2021 db/conf/icml/icml2021.html#NakagawaKS21 Atsushi Yaguchi Taiji Suzuki Shuhei Nitta Yukinobu Sakata Akiyuki Tanizawa Decomposable-Net: Scalable Low-Rank Compression for Neural Networks. 3249-3256 2021 IJCAI https://doi.org/10.24963/ijcai.2021/447 conf/ijcai/2021 db/conf/ijcai/ijcai2021.html#YaguchiSNST21 Taiji Suzuki Atsushi Nitanda Deep learning is adaptive to intrinsic dimensionality of model smoothness in anisotropic Besov space. 3609-3621 2021 NeurIPS https://proceedings.neurips.cc/paper/2021/hash/1dacb10f0623c67cb7dbb37587d8b38a-Abstract.html conf/nips/2021 db/conf/nips/neurips2021.html#SuzukiN21 Stefano Massaroli Michael Poli Sho Sonoda Taiji Suzuki Jinkyoo Park Atsushi Yamashita Hajime Asama Differentiable Multiple Shooting Layers. 16532-16544 2021 NeurIPS https://proceedings.neurips.cc/paper/2021/hash/89b9c689a57b82e59074c6ba09aa394d-Abstract.html conf/nips/2021 db/conf/nips/neurips2021.html#MassaroliPSSPYA21 Atsushi Nitanda Denny Wu Taiji Suzuki Particle Dual Averaging: Optimization of Mean Field Neural Network with Global Convergence Rate Analysis. 19608-19621 2021 NeurIPS https://proceedings.neurips.cc/paper/2021/hash/a34e1ddbb4d329167f50992ba59fe45a-Abstract.html conf/nips/2021 db/conf/nips/neurips2021.html#NitandaWS21 Chihiro Watanabe Taiji Suzuki AutoLL: Automatic Linear Layout of Graphs based on Deep Neural Network. 1-10 2021 SSCI https://doi.org/10.1109/SSCI50451.2021.9659893 conf/ssci/2021 db/conf/ssci/ssci2021.html#WatanabeS21

Tomoya Murata Taiji Suzuki Bias-Variance Reduced Local SGD for Less Heterogeneous Federated Learning. 2021 abs/2102.03198 CoRR https://arxiv.org/abs/2102.03198 db/journals/corr/corr2102.html#abs-2102-03198

Chihiro Watanabe Taiji Suzuki Goodness-of-fit Test on the Number of Biclusters in Relational Data Matrix. 2021 abs/2102.11658 CoRR https://arxiv.org/abs/2102.11658 db/journals/corr/corr2102.html#abs-2102-11658

Chihiro Watanabe Taiji Suzuki Deep Two-way Matrix Reordering for Relational Data Analysis. 2021 abs/2103.14203 CoRR https://arxiv.org/abs/2103.14203 db/journals/corr/corr2103.html#abs-2103-14203

Stefano Massaroli Michael Poli Sho Sonoda Taiji Suzuki Jinkyoo Park Atsushi Yamashita Hajime Asama Differentiable Multiple Shooting Layers. 2021 abs/2106.03885 CoRR https://arxiv.org/abs/2106.03885 db/journals/corr/corr2106.html#abs-2106-03885

Shunta Akiyama Taiji Suzuki On Learnability via Gradient Method for Two-Layer ReLU Neural Networks in Teacher-Student Setting. 2021 abs/2106.06251 CoRR https://arxiv.org/abs/2106.06251 db/journals/corr/corr2106.html#abs-2106-06251

Chihiro Watanabe Taiji Suzuki AutoLL: Automatic Linear Layout of Graphs based on Deep Neural Network. 2021 abs/2108.02431 CoRR https://arxiv.org/abs/2108.02431 db/journals/corr/corr2108.html#abs-2108-02431

Kishan Wimalawarne Taiji Suzuki Adaptive and Interpretable Graph Convolution Networks Using Generalized Pagerank. 2021 abs/2108.10636 CoRR https://arxiv.org/abs/2108.10636 db/journals/corr/corr2108.html#abs-2108-10636

Hiroaki Mikami Kenji Fukumizu Shogo Murai Shuji Suzuki Yuta Kikuchi Taiji Suzuki Shin-ichi Maeda Kohei Hayashi A Scaling Law for Synthetic-to-Real Transfer: A Measure of Pre-Training. 2021 abs/2108.11018 CoRR https://arxiv.org/abs/2108.11018 db/journals/corr/corr2108.html#abs-2108-11018

Hiroaki Kingetsu Kenichi Kobayashi 0001 Taiji Suzuki Neural Network Module Decomposition and Recomposition. 2021 abs/2112.13208 CoRR https://arxiv.org/abs/2112.13208 db/journals/corr/corr2112.html#abs-2112-13208

Shaogao Lv Zengyan Fan Heng Lian 0002 Taiji Suzuki Kenji Fukumizu A reproducing kernel Hilbert space approach to high dimensional partially varying coefficient model. 107039 2020 152 Comput. Stat. Data Anal. https://doi.org/10.1016/j.csda.2020.107039 db/journals/csda/csda152.html#LvFLSF20

Masaaki Takada Taiji Suzuki Hironori Fujisawa Independently Interpretable Lasso for Generalized Linear Models. 1168-1221 2020 32 Neural Comput. 6 https://doi.org/10.1162/neco_a_01279 https://www.wikidata.org/entity/Q94445434 db/journals/neco/neco32.html#TakadaSF20

Satoshi Hayakawa Taiji Suzuki On the minimax optimality and superiority of deep neural network learning over sparse parameter spaces. 343-361 2020 123 Neural Networks https://doi.org/10.1016/j.neunet.2019.12.014 https://www.wikidata.org/entity/Q92414883 db/journals/nn/nn123.html#HayakawaS20

Jingling Li Yanchao Sun Jiahao Su Taiji Suzuki Furong Huang Understanding Generalization in Deep Learning via Tensor Methods. 504-515 2020 AISTATS http://proceedings.mlr.press/v108/li20c.html conf/aistats/2020 db/conf/aistats/aistats2020.html#LiSSSH20 Atsushi Nitanda Taiji Suzuki Functional Gradient Boosting for Learning Residual-like Networks with Statistical Guarantees. 2981-2991 2020 AISTATS http://proceedings.mlr.press/v108/nitanda20a.html conf/aistats/2020 db/conf/aistats/aistats2020.html#NitandaS20 Laurent Dillard Yosuke Shinya Taiji Suzuki Domain Adaptation Regularization for Spectral Pruning. 2020 BMVC https://www.bmvc2020-conference.com/assets/papers/0453.pdf conf/bmvc/2020 db/conf/bmvc/bmvc2020.html#DillardSS20 Jimmy Ba Murat A. Erdogdu Taiji Suzuki Denny Wu Tianzong Zhang Generalization of Two-layer Neural Networks: An Asymptotic Viewpoint. 2020 ICLR https://openreview.net/forum?id=H1gBsgBYwH conf/iclr/2020 db/conf/iclr/iclr2020.html#BaESWZ20 Kenta Oono Taiji Suzuki Graph Neural Networks Exponentially Lose Expressive Power for Node Classification. 2020 ICLR https://openreview.net/forum?id=S1ldO2EFPr conf/iclr/2020 db/conf/iclr/iclr2020.html#OonoS20 Taiji Suzuki Hiroshi Abe Tomoaki Nishimura Compression based bound for non-compressed network: unified generalization error analysis of large compressible deep neural network. 2020 ICLR https://openreview.net/forum?id=ByeGzlrKwH conf/iclr/2020 db/conf/iclr/iclr2020.html#SuzukiAN20 Taiji Suzuki Hiroshi Abe Tomoya Murata Shingo Horiuchi Kotaro Ito Tokuma Wachi So Hirai Masatoshi Yukishima Tomoaki Nishimura Spectral Pruning: Compressing Deep Neural Networks via Spectral Analysis and its Generalization Error. 2839-2846 2020 IJCAI https://doi.org/10.24963/ijcai.2020/393 conf/ijcai/2020 db/conf/ijcai/ijcai2020.html#SuzukiAMHIWHYN20 Kenta Oono Taiji Suzuki Optimization and Generalization Analysis of Transduction through Gradient Boosting and Application to Multi-scale Graph Neural Networks. 2020 NeurIPS https://proceedings.neurips.cc/paper/2020/hash/dab49080d80c724aad5ebf158d63df41-Abstract.html conf/nips/2020 db/conf/nips/neurips2020.html#OonoS20 Taiji Suzuki Generalization bound of globally optimal non-convex neural network training: Transportation map estimation by infinite dimensional Langevin dynamics. 2020 NeurIPS https://proceedings.neurips.cc/paper/2020/hash/df1a336b7e0b0cb186de6e66800c43a9-Abstract.html conf/nips/2020 db/conf/nips/neurips2020.html#Suzuki20

Jingling Li Yanchao Sun Jiahao Su Taiji Suzuki Furong Huang Understanding Generalization in Deep Learning via Tensor Methods. 2020 abs/2001.05070 CoRR https://arxiv.org/abs/2001.05070 db/journals/corr/corr2001.html#abs-2001-05070

Boris Muzellec Kanji Sato Mathurin Massias Taiji Suzuki Dimension-free convergence rates for gradient Langevin dynamics in RKHS. 2020 abs/2003.00306 CoRR https://arxiv.org/abs/2003.00306 db/journals/corr/corr2003.html#abs-2003-00306

Yusuke Hayashi Taiji Suzuki Meta Cyclical Annealing Schedule: A Simple Approach to Avoiding Meta-Amortization Error. 2020 abs/2003.01889 CoRR https://arxiv.org/abs/2003.01889 db/journals/corr/corr2003.html#abs-2003-01889

Chihiro Watanabe Taiji Suzuki Selective Inference for Latent Block Models. 2020 abs/2005.13273 CoRR https://arxiv.org/abs/2005.13273 db/journals/corr/corr2005.html#abs-2005-13273

Kenta Oono Taiji Suzuki Optimization and Generalization Analysis of Transduction through Gradient Boosting and Application to Multi-scale Graph Neural Networks. 2020 abs/2006.08550 CoRR https://arxiv.org/abs/2006.08550 db/journals/corr/corr2006.html#abs-2006-08550

Shun-ichi Amari Jimmy Ba Roger B. Grosse Xuechen Li 0005 Atsushi Nitanda Taiji Suzuki Denny Wu Ji Xu When Does Preconditioning Help or Hurt Generalization? 2020 abs/2006.10732 CoRR https://arxiv.org/abs/2006.10732 db/journals/corr/corr2006.html#abs-2006-10732

Tomoya Murata Taiji Suzuki Gradient Descent in RKHS with Importance Labeling. 2020 abs/2006.10925 CoRR https://arxiv.org/abs/2006.10925 db/journals/corr/corr2006.html#abs-2006-10925

Atsushi Nitanda Taiji Suzuki Optimal Rates for Averaged Stochastic Gradient Descent under Neural Tangent Kernel Regime. 2020 abs/2006.12297 CoRR https://arxiv.org/abs/2006.12297 db/journals/corr/corr2006.html#abs-2006-12297

Taiji Suzuki Generalization bound of globally optimal non-convex neural network training: Transportation map estimation by infinite dimensional Langevin dynamics. 2020 abs/2007.05824 CoRR https://arxiv.org/abs/2007.05824 db/journals/corr/corr2007.html#abs-2007-05824

Kengo Machida Kuniaki Uto Koichi Shinoda Taiji Suzuki Neural Architecture Search Using Stable Rank of Convolutional Layers. 2020 abs/2009.09209 CoRR https://arxiv.org/abs/2009.09209 db/journals/corr/corr2009.html#abs-2009-09209

Kazuma Tsuji Taiji Suzuki Estimation error analysis of deep learning on the regression problem on the variable exponent Besov space. 2020 abs/2009.11285 CoRR https://arxiv.org/abs/2009.11285 db/journals/corr/corr2009.html#abs-2009-11285

Taiji Suzuki Shunta Akiyama Benefit of deep learning with non-convex noisy gradient descent: Provable excess risk bound and superiority to kernel methods. 2020 abs/2012.03224 CoRR https://arxiv.org/abs/2012.03224 db/journals/corr/corr2012.html#abs-2012-03224

Atsushi Nitanda Denny Wu Taiji Suzuki Particle Dual Averaging: Optimization of Mean Field Neural Networks with Global Convergence Rate Analysis. 2020 abs/2012.15477 CoRR https://arxiv.org/abs/2012.15477 db/journals/corr/corr2012.html#abs-2012-15477

Atsushi Nitanda Taiji Suzuki Stochastic Gradient Descent with Exponential Convergence Rates of Expected Classification Errors. 1417-1426 2019 AISTATS http://proceedings.mlr.press/v89/nitanda19a.html conf/aistats/2019 db/conf/aistats/aistats2019.html#NitandaS19 Heishiro Kanagawa Hayato Kobayashi Nobuyuki Shimizu Yukihiro Tagami Taiji Suzuki Cross-Domain Recommendation via Deep Domain Adaptation. 20-29 2019 ECIR (2) https://doi.org/10.1007/978-3-030-15719-7_3 conf/ecir/2019-2 db/conf/ecir/ecir2019-2.html#KanagawaKSTS19 Yosuke Shinya Edgar Simo-Serra Taiji Suzuki Understanding the Effects of Pre-Training for Object Detectors via Eigenspectrum. 1931-1941 2019 ICCV Workshops https://doi.org/10.1109/ICCVW.2019.00242 conf/iccvw/2019 db/conf/iccvw/iccvw2019.html#ShinyaSS19 Atsushi Nitanda Tomoya Murata Taiji Suzuki Sharp Characterization of Optimal Minibatch Size for Stochastic Finite Sum Convex Optimization. 488-497 2019 ICDM https://doi.org/10.1109/ICDM.2019.00059 conf/icdm/2019 db/conf/icdm/icdm2019.html#NitandaMS19 Taiji Suzuki Adaptivity of deep ReLU network for learning in Besov and mixed smooth Besov spaces: optimal rate and curse of dimensionality. 2019 ICLR (Poster) https://openreview.net/forum?id=H1ebTsActm conf/iclr/2019 db/conf/iclr/iclr2019.html#Suzuki19 Kenta Oono Taiji Suzuki Approximation and non-parametric estimation of ResNet-type convolutional neural networks. 4922-4931 2019 ICML http://proceedings.mlr.press/v97/oono19a.html conf/icml/2019 db/conf/icml/icml2019.html#OonoS19 Wee Sun Lee Taiji Suzuki Proceedings of The 11th Asian Conference on Machine Learning, ACML 2019, 17-19 November 2019, Nagoya, Japan. ACML PMLR Proceedings of Machine Learning Research 101 2019 http://proceedings.mlr.press/v101/ db/conf/acml/acml2019.html

Kenta Oono Taiji Suzuki Approximation and Non-parametric Estimation of ResNet-type Convolutional Neural Networks. 2019 abs/1903.10047 CoRR http://arxiv.org/abs/1903.10047 db/journals/corr/corr1903.html#abs-1903-10047

Satoshi Hayakawa Taiji Suzuki On the minimax optimality and superiority of deep neural network learning over sparse parameter spaces. 2019 abs/1905.09195 CoRR http://arxiv.org/abs/1905.09195 db/journals/corr/corr1905.html#abs-1905-09195

Atsushi Nitanda Taiji Suzuki Refined Generalization Analysis of Gradient Descent for Over-parameterized Two-layer Neural Networks with Smooth Activations on Classification Problems. 2019 abs/1905.09870 CoRR http://arxiv.org/abs/1905.09870 db/journals/corr/corr1905.html#abs-1905-09870

Kenta Oono Taiji Suzuki On Asymptotic Behaviors of Graph CNNs from Dynamical Systems Perspective. 2019 abs/1905.10947 CoRR http://arxiv.org/abs/1905.10947 db/journals/corr/corr1905.html#abs-1905-10947

Tomoya Murata Taiji Suzuki Accelerated Sparsified SGD with Error Feedback. 2019 abs/1905.12224 CoRR http://arxiv.org/abs/1905.12224 db/journals/corr/corr1905.html#abs-1905-12224

Chihiro Watanabe Taiji Suzuki Goodness-of-fit Test for Latent Block Models. 2019 abs/1906.03886 CoRR http://arxiv.org/abs/1906.03886 db/journals/corr/corr1906.html#abs-1906-03886

Kosuke Haruki Taiji Suzuki Yohei Hamakawa Takeshi Toda Ryuji Sakai Masahiro Ozawa Mitsuhiro Kimura Gradient Noise Convolution (GNC): Smoothing Loss Function for Distributed Large-Batch SGD. 2019 abs/1906.10822 CoRR http://arxiv.org/abs/1906.10822 db/journals/corr/corr1906.html#abs-1906-10822

Yosuke Shinya Edgar Simo-Serra Taiji Suzuki Understanding the Effects of Pre-Training for Object Detectors via Eigenspectrum. 2019 abs/1909.04021 CoRR http://arxiv.org/abs/1909.04021 db/journals/corr/corr1909.html#abs-1909-04021

Taiji Suzuki Compression based bound for non-compressed network: unified generalization error analysis of large compressible deep neural network. 2019 abs/1909.11274 CoRR http://arxiv.org/abs/1909.11274 db/journals/corr/corr1909.html#abs-1909-11274

Taiji Suzuki Atsushi Nitanda Deep learning is adaptive to intrinsic dimensionality of model smoothness in anisotropic Besov space. 2019 abs/1910.12799 CoRR http://arxiv.org/abs/1910.12799 db/journals/corr/corr1910.html#abs-1910-12799

Atsushi Yaguchi Taiji Suzuki Shuhei Nitta Yukinobu Sakata Akiyuki Tanizawa Scalable Deep Neural Networks via Low-Rank Matrix Factorization. 2019 abs/1910.13141 CoRR http://arxiv.org/abs/1910.13141 db/journals/corr/corr1910.html#abs-1910-13141

Shingo Yashima Atsushi Nitanda Taiji Suzuki Exponential Convergence Rates of Classification Errors on Learning with SGD and Random Features. 2019 abs/1911.05350 CoRR http://arxiv.org/abs/1911.05350 db/journals/corr/corr1911.html#abs-1911-05350

Laurent Dillard Yosuke Shinya Taiji Suzuki Domain Adaptation Regularization for Spectral Pruning. 2019 abs/1912.11853 CoRR http://arxiv.org/abs/1912.11853 db/journals/corr/corr1912.html#abs-1912-11853

Yuichi Mori Taiji Suzuki Generalized ridge estimator and model selection criteria in multivariate linear regression. 243-261 2018 165 J. Multivar. Anal. https://doi.org/10.1016/j.jmva.2017.12.006 db/journals/ma/ma165.html#MoriS18

Masaaki Takada Taiji Suzuki Hironori Fujisawa Independently Interpretable Lasso: A New Regularizer for Sparse Regression with Uncorrelated Variables. 454-463 2018 AISTATS http://proceedings.mlr.press/v84/takada18a.html conf/aistats/2018 db/conf/aistats/aistats2018.html#TakadaSF18 Atsushi Nitanda Taiji Suzuki Gradient Layer: Enhancing the Convergence of Adversarial Training for Generative Models. 1008-1016 2018 AISTATS http://proceedings.mlr.press/v84/nitanda18a.html conf/aistats/2018 db/conf/aistats/aistats2018.html#NitandaS18 Taiji Suzuki Fast generalization error bound of deep learning from a kernel perspective. 1397-1406 2018 AISTATS http://proceedings.mlr.press/v84/suzuki18a.html conf/aistats/2018 db/conf/aistats/aistats2018.html#Suzuki18 Kazuo Yonekura Hitoshi Hattori Taiji Suzuki Short-term local weather forecast using dense weather station by deep neural network. 1683-1690 2018 IEEE BigData https://doi.org/10.1109/BigData.2018.8622195 conf/bigdataconf/2018 db/conf/bigdataconf/bigdataconf2018.html#YonekuraHS18 Atsushi Nitanda Taiji Suzuki Functional Gradient Boosting based on Residual Network Perception. 3816-3825 2018 ICML http://proceedings.mlr.press/v80/nitanda18a.html conf/icml/2018 db/conf/icml/icml2018.html#NitandaS18 Atsushi Yaguchi Taiji Suzuki Wataru Asano Shuhei Nitta Yukinobu Sakata Akiyuki Tanizawa Adam Induces Implicit Weight Sparsity in Rectifier Neural Networks. 318-325 2018 ICMLA https://doi.org/10.1109/ICMLA.2018.00054 conf/icmla/2018 db/conf/icmla/icmla2018.html#YaguchiSANST18 Tomoya Murata Taiji Suzuki Sample Efficient Stochastic Gradient Iterative Hard Thresholding Method for Stochastic Sparse Linear Regression with Limited Attribute Observation. 5317-5326 2018 NeurIPS https://proceedings.neurips.cc/paper/2018/hash/c3a690be93aa602ee2dc0ccab5b7b67e-Abstract.html http://papers.nips.cc/paper/7777-sample-efficient-stochastic-gradient-iterative-hard-thresholding-method-for-stochastic-sparse-linear-regression-with-limited-attribute-observation conf/nips/2018 db/conf/nips/nips2018.html#MurataS18

Atsushi Nitanda Taiji Suzuki Gradient Layer: Enhancing the Convergence of Adversarial Training for Generative Models. 2018 abs/1801.02227 CoRR http://arxiv.org/abs/1801.02227 db/journals/corr/corr1801.html#abs-1801-02227

Atsushi Nitanda Taiji Suzuki Functional Gradient Boosting based on Residual Network Perception. 2018 abs/1802.09031 CoRR http://arxiv.org/abs/1802.09031 db/journals/corr/corr1802.html#abs-1802-09031

Heishiro Kanagawa Hayato Kobayashi Nobuyuki Shimizu Yukihiro Tagami Taiji Suzuki Cross-domain Recommendation via Deep Domain Adaptation. 2018 abs/1803.03018 CoRR http://arxiv.org/abs/1803.03018 db/journals/corr/corr1803.html#abs-1803-03018

Atsushi Nitanda Taiji Suzuki Stochastic Gradient Descent with Exponential Convergence Rates of Expected Classification Errors. 2018 abs/1806.05438 CoRR http://arxiv.org/abs/1806.05438 db/journals/corr/corr1806.html#abs-1806-05438

Taiji Suzuki Hiroshi Abe Tomoya Murata Shingo Horiuchi Kotaro Ito Tokuma Wachi So Hirai Masatoshi Yukishima Tomoaki Nishimura Spectral-Pruning: Compressing deep neural network via spectral analysis. 2018 abs/1808.08558 CoRR http://arxiv.org/abs/1808.08558 db/journals/corr/corr1808.html#abs-1808-08558

Tomoya Murata Taiji Suzuki Sample Efficient Stochastic Gradient Iterative Hard Thresholding Method for Stochastic Sparse Linear Regression with Limited Attribute Observation. 2018 abs/1809.01765 CoRR http://arxiv.org/abs/1809.01765 db/journals/corr/corr1809.html#abs-1809-01765

Taiji Suzuki Adaptivity of deep ReLU network for learning in Besov and mixed smooth Besov spaces: optimal rate and curse of dimensionality. 2018 abs/1810.08033 CoRR http://arxiv.org/abs/1810.08033 db/journals/corr/corr1810.html#abs-1810-08033

Atsushi Yaguchi Taiji Suzuki Wataru Asano Shuhei Nitta Yukinobu Sakata Akiyuki Tanizawa Adam Induces Implicit Weight Sparsity in Rectifier Neural Networks. 2018 abs/1812.08119 CoRR http://arxiv.org/abs/1812.08119 db/journals/corr/corr1812.html#abs-1812-08119

Atsushi Nitanda Taiji Suzuki Stochastic Difference of Convex Algorithm and its Application to Training Deep Boltzmann Machines. 470-478 2017 AISTATS http://proceedings.mlr.press/v54/nitanda17a.html conf/aistats/2017 db/conf/aistats/aistats2017.html#NitandaS17 Tomoya Murata Taiji Suzuki Doubly Accelerated Stochastic Variance Reduced Dual Averaging Method for Regularized Empirical Risk Minimization. 608-617 2017 NIPS https://proceedings.neurips.cc/paper/2017/hash/75fc093c0ee742f6dddaa13fff98f104-Abstract.html http://papers.nips.cc/paper/6663-doubly-accelerated-stochastic-variance-reduced-dual-averaging-method-for-regularized-empirical-risk-minimization conf/nips/2017 db/conf/nips/nips2017.html#MurataS17 Song Liu 0002 Akiko Takeda Taiji Suzuki Kenji Fukumizu Trimmed Density Ratio Estimation. 4518-4528 2017 NIPS https://proceedings.neurips.cc/paper/2017/hash/ea204361fe7f024b130143eb3e189a18-Abstract.html http://papers.nips.cc/paper/7038-trimmed-density-ratio-estimation conf/nips/2017 db/conf/nips/nips2017.html#LiuTSF17

Tomoya Murata Taiji Suzuki Doubly Accelerated Stochastic Variance Reduced Dual Averaging Method for Regularized Empirical Risk Minimization. 2017 abs/1703.00439 CoRR http://arxiv.org/abs/1703.00439 db/journals/corr/corr1703.html#MurataS17

Taiji Suzuki Fast learning rate of deep learning via a kernel perspective. 2017 abs/1705.10182 CoRR http://arxiv.org/abs/1705.10182 db/journals/corr/corr1705.html#Suzuki17

Atsushi Nitanda Taiji Suzuki Stochastic Particle Gradient Descent for Infinite Ensembles. 2017 abs/1712.05438 CoRR http://arxiv.org/abs/1712.05438 db/journals/corr/corr1712.html#abs-1712-05438

Yoshito Hirata Kai Morino Taiji Suzuki Qian Guo 0002 Hiroshi Fukuhara Kazuyuki Aihara System identification and parameter estimation in mathematical medicine: examples demonstrated for prostate cancer. 13-19 2016 4 Quant. Biol. 1 https://doi.org/10.1007/s40484-016-0059-0 db/journals/quanbio/quanbio4.html#HirataMSGFA16

Song Liu 0002 Taiji Suzuki Masashi Sugiyama Kenji Fukumizu Structure Learning of Partitioned Markov Networks. 439-448 2016 ICML http://proceedings.mlr.press/v48/liuc16.html conf/icml/2016 db/conf/icml/icml2016.html#LiuSSF16 Heishiro Kanagawa Taiji Suzuki Hayato Kobayashi Nobuyuki Shimizu Yukihiro Tagami Gaussian process nonparametric tensor estimator and its minimax optimality. 1632-1641 2016 ICML http://proceedings.mlr.press/v48/kanagawa16.html conf/icml/2016 db/conf/icml/icml2016.html#KanagawaSKST16 Taiji Suzuki Heishiro Kanagawa Hayato Kobayashi Nobuyuki Shimizu Yukihiro Tagami Minimax Optimal Alternating Minimization for Kernel Nonparametric Tensor Learning. 3783-3791 2016 NIPS https://proceedings.neurips.cc/paper/2016/hash/b4568df26077653eeadf29596708c94b-Abstract.html http://papers.nips.cc/paper/6419-minimax-optimal-alternating-minimization-for-kernel-nonparametric-tensor-learning conf/nips/2016 db/conf/nips/nips2016.html#SuzukiKKST16

Tomoya Murata Taiji Suzuki Stochastic dual averaging methods using variance reduction techniques for regularized empirical risk minimization problems. 2016 abs/1603.02412 CoRR http://arxiv.org/abs/1603.02412 db/journals/corr/corr1603.html#MurataS16

Song Liu 0002 Taiji Suzuki Masashi Sugiyama Support Consistency of Direct Sparse-Change Learning in Markov Networks. 2785-2791 2015 AAAI https://doi.org/10.1609/aaai.v29i1.9566 https://www.wikidata.org/entity/Q114887031 conf/aaai/2015 db/conf/aaai/aaai2015.html#LiuSS15 Satoshi Hara 0001 Tetsuro Morimura Toshihiro Takahashi Hiroki Yanagisawa Taiji Suzuki A Consistent Method for Graph Based Anomaly Localization. 2015 AISTATS http://proceedings.mlr.press/v38/hara15.html conf/aistats/2015 db/conf/aistats/aistats2015.html#HaraMTYS15 Taiji Suzuki Convergence rate of Bayesian tensor estimator and its minimax optimality. 1273-1282 2015 ICML http://proceedings.mlr.press/v37/suzuki15.html conf/icml/2015 db/conf/icml/icml2015.html#Suzuki15

Song Liu 0002 John A. Quinn Michael U. Gutmann Taiji Suzuki Masashi Sugiyama Direct Learning of Sparse Changes in Markov Networks by Density Ratio Estimation. 1169-1197 2014 26 Neural Comput. 6 https://doi.org/10.1162/NECO_a_00589 https://www.wikidata.org/entity/Q43483663 db/journals/neco/neco26.html#LiuQGSS14

Taiji Suzuki Stochastic Dual Coordinate Ascent with Alternating Direction Method of Multipliers. 736-744 2014 ICML http://proceedings.mlr.press/v32/suzuki14.html conf/icml/2014 db/conf/icml/icml2014.html#Suzuki14

Taiji Suzuki Convergence rate of Bayesian tensor estimator: Optimal rate without restricted strong convexity. 2014 abs/1408.3092 CoRR http://arxiv.org/abs/1408.3092 db/journals/corr/corr1408.html#Suzuki14a

Masashi Sugiyama Song Liu 0002 Marthinus Christoffel du Plessis Masao Yamanaka Makoto Yamada Taiji Suzuki Takafumi Kanamori Direct Divergence Approximation between Probability Distributions and Its Applications in Machine Learning. 99-111 2013 7 J. Comput. Sci. Eng. 2 https://doi.org/10.5626/JCSE.2013.7.2.99 https://www.wikidata.org/entity/Q114880708 db/journals/jcse/jcse7.html#SugiyamaLPYYSK13

Takafumi Kanamori Akiko Takeda Taiji Suzuki Conjugate relation between loss functions and uncertainty sets in classification problems. 1461-1504 2013 14 J. Mach. Learn. Res. 1 https://dl.acm.org/doi/10.5555/2567709.2567710 db/journals/jmlr/jmlr14.html#KanamoriTS13

Taiji Suzuki Improvement of multiple kernel learning using adaptively weighted regularization. 49-52 2013 5 JSIAM Lett. https://doi.org/10.14495/jsiaml.5.49 db/journals/jsiaml/jsiaml5.html#Suzuki13

Takafumi Kanamori Taiji Suzuki Masashi Sugiyama Computational complexity of kernel-based density-ratio estimation: a condition number analysis. 431-460 2013 90 Mach. Learn. 3 https://doi.org/10.1007/s10994-012-5323-6 db/journals/ml/ml90.html#KanamoriSS13

Taiji Suzuki Masashi Sugiyama Sufficient Dimension Reduction via Squared-Loss Mutual Information Estimation. 725-758 2013 25 Neural Comput. 3 https://doi.org/10.1162/NECO_a_00407 https://www.wikidata.org/entity/Q50767941 db/journals/neco/neco25.html#SuzukiS13

Makoto Yamada Taiji Suzuki Takafumi Kanamori Hirotaka Hachiya Masashi Sugiyama Relative Density-Ratio Estimation for Robust Distribution Comparison. 1324-1370 2013 25 Neural Comput. 5 https://doi.org/10.1162/NECO_a_00442 https://www.wikidata.org/entity/Q44657040 db/journals/neco/neco25.html#YamadaSKHS13

Masashi Sugiyama Takafumi Kanamori Taiji Suzuki Marthinus Christoffel du Plessis Song Liu 0002 Ichiro Takeuchi Density-Difference Estimation. 2734-2775 2013 25 Neural Comput. 10 https://doi.org/10.1162/NECO_a_00492 https://www.wikidata.org/entity/Q44830143 db/journals/neco/neco25.html#SugiyamaKSPLT13

Taiji Suzuki Dual Averaging and Proximal Gradient Descent for Online Alternating Direction Multiplier Method. 392-400 2013 ICML (1) http://proceedings.mlr.press/v28/suzuki13.html conf/icml/2013 db/conf/icml/icml2013.html#Suzuki13 Ryota Tomioka Taiji Suzuki Convex Tensor Decomposition via Structured Schatten Norm Regularization. 1331-1339 2013 NIPS https://proceedings.neurips.cc/paper/2013/hash/109a0ca3bc27f3e96597370d5c8cf03d-Abstract.html http://papers.nips.cc/paper/4985-convex-tensor-decomposition-via-structured-schatten-norm-regularization conf/nips/2013 db/conf/nips/nips2013.html#TomiokaS13

Ryota Tomioka Taiji Suzuki Convex Tensor Decomposition via Structured Schatten Norm Regularization http://arxiv.org/abs/1303.6370 2013 CoRR abs/1303.6370 db/journals/corr/corr1303.html#abs-1303-6370

Masashi Sugiyama Taiji Suzuki Takafumi Kanamori Density Ratio Estimation in Machine Learning. I-XII, 1-329 Cambridge University Press 2012 978-0-521-19017-6 http://www.cambridge.org/de/academic/subjects/computer-science/pattern-recognition-and-machine-learning/density-ratio-estimation-machine-learning?format=HB

Takafumi Kanamori Taiji Suzuki Masashi Sugiyama Statistical analysis of kernel-based least-squares density-ratio estimation. 335-367 2012 86 Mach. Learn. 3 https://doi.org/10.1007/s10994-011-5266-3 db/journals/ml/ml86.html#KanamoriSS12

Takafumi Kanamori Taiji Suzuki Masashi Sugiyama f-Divergence Estimation and Two-Sample Homogeneity Test Under Semiparametric Density-Ratio Models. 708-720 2012 58 IEEE Trans. Inf. Theory 2 https://doi.org/10.1109/TIT.2011.2163380 db/journals/tit/tit58.html#KanamoriSS12

Taiji Suzuki PAC-Bayesian Bound for Gaussian Process Regression and Multiple Kernel Additive Model. 8.1-8.20 2012 COLT http://proceedings.mlr.press/v23/suzuki12/suzuki12.pdf conf/colt/2012 db/journals/jmlr/jmlrp23.html#Suzuki12 Masashi Sugiyama Takafumi Kanamori Taiji Suzuki Marthinus Christoffel du Plessis Song Liu 0002 Ichiro Takeuchi Density-Difference Estimation. 692-700 2012 NIPS https://proceedings.neurips.cc/paper/2012/hash/f2fc990265c712c49d51a18a32b39f0c-Abstract.html http://papers.nips.cc/paper/4641-density-difference-estimation conf/nips/2012 db/conf/nips/nips2012.html#SugiyamaKSPLT12 Takafumi Kanamori Akiko Takeda Taiji Suzuki A Conjugate Property between Loss Functions and Uncertainty Sets in Classification Problems. 29.1-29.23 COLT 2012 conf/colt/2012 http://proceedings.mlr.press/v23/kanamori12/kanamori12.pdf db/journals/jmlr/jmlrp23.html#KanamoriTS12 Taiji Suzuki Masashi Sugiyama Fast Learning Rate of Multiple Kernel Learning: Trade-Off between Sparsity and Smoothness. 1152-1183 AISTATS 2012 conf/aistats/2012 http://proceedings.mlr.press/v22/suzuki12.html db/journals/jmlr/jmlrp22.html#SuzukiS12

Takafumi Kanamori Akiko Takeda Taiji Suzuki A Conjugate Property between Loss Functions and Uncertainty Sets in Classification Problems http://arxiv.org/abs/1204.6583 2012 CoRR abs/1204.6583 db/journals/corr/corr1204.html#abs-1204-6583

Masashi Sugiyama Takafumi Kanamori Taiji Suzuki Marthinus Christoffel du Plessis Song Liu 0002 Ichiro Takeuchi Density-Difference Estimation http://arxiv.org/abs/1207.0099 2012 CoRR abs/1207.0099 db/journals/corr/corr1207.html#abs-1207-0099

Masashi Sugiyama Taiji Suzuki Least-Squares Independence Test. 1333-1336 2011 94-D IEICE Trans. Inf. Syst. 6 https://doi.org/10.1587/transinf.E94.D.1333 http://search.ieice.org/bin/summary.php?id=e94-d_6_1333 db/journals/ieicet/ieicet94d.html#SugiyamaS11

Ryota Tomioka Taiji Suzuki Masashi Sugiyama Super-Linear Convergence of Dual Augmented Lagrangian Algorithm for Sparsity Regularized Estimation. 1537-1586 2011 12 J. Mach. Learn. Res. https://dl.acm.org/doi/10.5555/1953048.2021050 db/journals/jmlr/jmlr12.html#TomiokaSS11

Taiji Suzuki Ryota Tomioka SpicyMKL: a fast algorithm for Multiple Kernel Learning with thousands of kernels. 77-108 2011 85 Mach. Learn. 1-2 https://doi.org/10.1007/s10994-011-5252-9 db/journals/ml/ml85.html#SuzukiT11

Taiji Suzuki Masashi Sugiyama Least-Squares Independent Component Analysis. 284-301 2011 23 Neural Comput. 1 https://doi.org/10.1162/NECO_a_00062 https://www.wikidata.org/entity/Q51645461 db/journals/neco/neco23.html#SuzukiS11

Masashi Sugiyama Makoto Yamada Paul von Bünau Taiji Suzuki Takafumi Kanamori Motoaki Kawanabe Direct density-ratio estimation with dimensionality reduction via least-squares hetero-distributional subspace search. 183-198 2011 24 Neural Networks 2 https://doi.org/10.1016/j.neunet.2010.10.005 https://www.wikidata.org/entity/Q37808029 db/journals/nn/nn24.html#SugiyamaYBSKK11

Masashi Sugiyama Taiji Suzuki Yuta Itoh 0001 Takafumi Kanamori Manabu Kimura Least-squares two-sample test. 735-751 2011 24 Neural Networks 7 https://doi.org/10.1016/j.neunet.2011.04.003 https://www.wikidata.org/entity/Q51568967 db/journals/nn/nn24.html#SugiyamaSIKK11

Makoto Yamada Taiji Suzuki Takafumi Kanamori Hirotaka Hachiya Masashi Sugiyama Relative Density-Ratio Estimation for Robust Distribution Comparison. 594-602 2011 NIPS https://proceedings.neurips.cc/paper/2011/hash/d1f255a373a3cef72e03aa9d980c7eca-Abstract.html http://papers.nips.cc/paper/4254-relative-density-ratio-estimation-for-robust-distribution-comparison conf/nips/2011 db/conf/nips/nips2011.html#YamadaSKHS11 Ryota Tomioka Taiji Suzuki Kohei Hayashi Hisashi Kashima Statistical Performance of Convex Tensor Decomposition. 972-980 2011 NIPS https://proceedings.neurips.cc/paper/2011/hash/b2eeb7362ef83deff5c7813a67e14f0a-Abstract.html http://papers.nips.cc/paper/4453-statistical-performance-of-convex-tensor-decomposition conf/nips/2011 db/conf/nips/nips2011.html#TomiokaSHK11 Taiji Suzuki Unifying Framework for Fast Learning Rate of Non-Sparse Multiple Kernel Learning. 1575-1583 2011 NIPS https://proceedings.neurips.cc/paper/2011/hash/aa169b49b583a2b5af89203c2b78c67c-Abstract.html http://papers.nips.cc/paper/4200-unifying-framework-for-fast-learning-rate-of-non-sparse-multiple-kernel-learning conf/nips/2011 db/conf/nips/nips2011.html#Suzuki11

Masashi Sugiyama Ichiro Takeuchi Taiji Suzuki Takafumi Kanamori Hirotaka Hachiya Daisuke Okanohara Least-Squares Conditional Density Estimation. 583-594 2010 93-D IEICE Trans. Inf. Syst. 3 https://doi.org/10.1587/transinf.E93.D.583 http://search.ieice.org/bin/summary.php?id=e93-d_3_583 db/journals/ieicet/ieicet93d.html#SugiyamaTSKHO10

Takafumi Kanamori Taiji Suzuki Masashi Sugiyama Theoretical Analysis of Density Ratio Estimation. 787-798 2010 93-A IEICE Trans. Fundam. Electron. Commun. Comput. Sci. 4 https://doi.org/10.1587/transfun.E93.A.787 http://search.ieice.org/bin/summary.php?id=e93-a_4_787 db/journals/ieicet/ieicet93a.html#KanamoriSS10

Ryota Tomioka Taiji Suzuki Masashi Sugiyama Hisashi Kashima A Fast Augmented Lagrangian Algorithm for Learning Low-Rank Matrices. 1087-1094 2010 ICML https://icml.cc/Conferences/2010/papers/556.pdf conf/icml/2010 db/conf/icml/icml2010.html#TomiokaSSK10 Masashi Sugiyama Satoshi Hara 0001 Paul von Bünau Taiji Suzuki Takafumi Kanamori Motoaki Kawanabe Direct Density Ratio Estimation with Dimensionality Reduction. 595-606 2010 SDM https://doi.org/10.1137/1.9781611972801.52 conf/sdm/2010 db/conf/sdm/sdm2010.html#SugiyamaHBSKK10 Masashi Sugiyama Ichiro Takeuchi Taiji Suzuki Takafumi Kanamori Hirotaka Hachiya Daisuke Okanohara Conditional Density Estimation via Least-Squares Density Ratio Estimation. 781-788 AISTATS 2010 conf/aistats/2010 http://proceedings.mlr.press/v9/sugiyama10a.html db/journals/jmlr/jmlrp9.html#SugiyamaTSKHO10 Taiji Suzuki Masashi Sugiyama Sufficient Dimension Reduction via Squared-loss Mutual Information Estimation. 804-811 AISTATS 2010 conf/aistats/2010 http://proceedings.mlr.press/v9/suzuki10a.html db/journals/jmlr/jmlrp9.html#SuzukiS10

Ryota Tomioka Taiji Suzuki Regularization Strategies and Empirical Bayesian Learning for MKL. 2010 abs/1011.3090 CoRR http://arxiv.org/abs/1011.3090 db/journals/corr/corr1011.html#TomiokaS10

Taiji Suzuki Masashi Sugiyama Takafumi Kanamori Jun Sese Mutual information estimation reveals global associations between stimuli and biological processes. 2009 10 BMC Bioinform. S-1 https://doi.org/10.1186/1471-2105-10-S1-S52 https://www.wikidata.org/entity/Q33407884 db/journals/bmcbi/bmcbi10S.html#SuzukiSKS09

Masashi Sugiyama Takafumi Kanamori Taiji Suzuki Shohei Hido Jun Sese Ichiro Takeuchi Liwei Wang 0001 A Density-ratio Framework for Statistical Data Processing. 962-987 2009 4 Inf. Media Technol. 4 https://doi.org/10.11185/imt.4.962 db/journals/imt/imt4.html#SugiyamaKSHST009

Masashi Sugiyama Takafumi Kanamori Taiji Suzuki Shohei Hido Jun Sese Ichiro Takeuchi Liwei Wang 0001 A Density-ratio Framework for Statistical Data Processing. 183-208 2009 1 IPSJ Trans. Comput. Vis. Appl. https://doi.org/10.2197/ipsjtcva.1.183 db/journals/ipsjtcva/ipsjtcva1.html#SugiyamaKSHSTW09

Taiji Suzuki Masashi Sugiyama Estimating Squared-Loss Mutual Information for Independent Component Analysis. 130-137 2009 ICA https://doi.org/10.1007/978-3-642-00599-2_17 conf/ida/2009 db/conf/ida/ida2009.html#SuzukiS09 Taiji Suzuki Masashi Sugiyama Toshiyuki Tanaka 0003 Mutual information approximation via maximum likelihood estimation of density ratio. 463-467 2009 ISIT https://doi.org/10.1109/ISIT.2009.5205712 conf/isit/2009 db/conf/isit/isit2009.html#SuzukiST09

Ryota Tomioka Taiji Suzuki Masashi Sugiyama Super-Linear Convergence of Dual Augmented-Lagrangian Algorithm for Sparsity Regularized Estimation. 2009 abs/0911.4046 CoRR http://arxiv.org/abs/0911.4046 db/journals/corr/corr0911.html#TomiokaSS09

Taiji Suzuki Masashi Sugiyama Jun Sese Takafumi Kanamori Approximating Mutual Information by Maximum Likelihood Density Ratio Estimation. 5-20 FSDM 2008 conf/pkdd/2008fsdm http://proceedings.mlr.press/v4/suzuki08a.html db/journals/jmlr/jmlrp4.html#SuzukiSSK08 Taiji Suzuki Takamasa Koshizen Kazuyuki Aihara Hiroshi Tsujino Learning to estimate user interest utilizing the variational Bayes estimator. 94-99 2005 conf/isda/2005 ISDA https://doi.org/10.1109/ISDA.2005.59 https://doi.ieeecomputersociety.org/10.1109/ISDA.2005.59 db/conf/isda/isda2005.html#SuzukiAKT05 Hiroshi Abe Kazuyuki Aihara Shunta Akiyama Shun-ichi Amari Qi An 0001 Hajime Asama Wataru Asano Ryoya Awano Jimmy Ba Mingyuan Bai Yoshua Bengio Alberto Bietti Mathieu Blondel Garyk Brixi Dake Bu Paul von Bünau Zhongyi Cai Xin Cao 0001 Yuan Cao 0006 Volkan Cevher Jianfei Chen 0001 Wuyang Chen 0001 Yihang Chen 0003 Yilan Chen 0002 Yongqiang Chen 0002 Zonghao Chen Ji Cheng 0001 Björn Deiseroth Laurent Dillard Murat A. Erdogdu Stefano Ermon Zengyan Fan Guoji Fu Hironori Fujisawa Hiroshi Fukuhara Kenji Fukumizu Marzyeh Ghassemi Raja Giryes Arthur Gretton Roger B. GrosseRoger Baker Grosse Qian Guo 0002 Michael U. Gutmann Hirotaka Hachiya Yohei Hamakawa Andi Han Satoshi Hara 0001 Kosuke Haruki Ramin M. Hasani Yuka Hashimoto Hitoshi Hattori Satoshi Hayakawa Kohei Hayashi Yusuke Hayashi Shohei Hido Brian L. Hie Rei Higuchi Ryusuke Hirai So Hirai Yoshito Hirata Motokiyo Hirayama Shingo Horiuchi Furong Huang Wei Huang 0034 Isao Ishikawa Kohta Ishikawa Noboru Isobe Kotaro Ito Yuta Itoh 0001 Jiarui Jiang Damian Tan Xing Kai Heishiro Kanagawa Takafumi Kanamori Hisashi Kashima Keizo Kato Takanobu Kawahara Motoaki Kawanabe Ryotaro Kawata Kristian Kersting Yuta Kikuchi Juno Kim Manabu Kimura Mitsuhiro Kimura Hiroaki Kingetsu Hirota Kinoshita Yuri Kinoshita Hayato Kobayashi Kenichi Kobayashi 0001 Sosuke Kobayashi Takamasa Koshizen Masanori Koyama Kento Kuwataka Chieh-Hsin Lai Sophie Langer Mathias Lechner Anzelle Lee Jason Lee Jason D. Lee Wee Sun Lee Bingrui Li Jingling Li Xuechen Li 0005 Zhu Li Heng Lian 0002 Fanghui Liu 0001 Song Liu 0002 Shaogao Lv Kengo Machida Shin-ichi Maeda Stefano Massaroli Mathurin Massias Kohsei Matsutani Arya Mazumdar Dimitri Meunier Hiroaki Mikami Yuki Mitsufuji Kohei Miyaguchi Yuichi Mori Tetsuro Morimura Kai Morino Alessandro Moro Alireza Mousavi-Hosseini Albert Rodríguez Mulet Shogo Murai Naoki Murata Tomoya Murata Boris Muzellec Hiroki Naganuma Akira Nakagawa Hideyuki Nakagawa Tai Nakamaki Eric Nguyen Liqiang Nie Naoki Nishikawa Tomoaki Nishimura Yuto Nishimura Atsushi Nitanda Shuhei Nitta Gang Niu 0001 Masahiro Nomura Junsoo Oh Daisuke Okanohara Kazusato Oko Sho Okumoto Kenta Oono Masahiro Ozawa Jinkyoo Park Rom N. Parnichkun Gabriel Peyré Marthinus Christoffel du Plessis Michael Poli Pragaash Ponnusamy John A. Quinn Christopher Ré Mizuki Sakaguchi Ryuji Sakai Yukinobu Sakata Michael E. SanderMichael Eli Sander Ikuro Sato Kanji Sato Taro Sawaki Dragos Secrieru Jun Sese Ye Shi 0001 Nobuyuki Shimizu Koichi Shinoda Yosuke Shinya Edgar Simo-Serra Jimmy T. H. Smith Yujin Song Sho Sonoda Jiahao Su Masashi Sugiyama Shengyang Sun Yanchao Sun Yuekai Sun Atsushi Suzuki 0002 Keita Suzuki Shuji Suzuki Yukihiro Tagami Masaaki Takada Toshihiro Takahashi Shokichi Takakura Akiko Takeda Nobuhito Takenouchi Ichiro Takeuchi Yuhta Takida Takahiro Takimoto Toshiyuki Tanaka 0003 Akiyuki Tanizawa Armin W. Thomas Feng Tian 0006 Takeshi Toda Seiya Tokui Ryota Tomioka Kazuma Tsuji Hiroshi Tsujino Neehal Tumma Kazuki Uematsu Toshimitsu Uesaka Kuniaki Uto Samuel Vaiter Akifumi Wachi Tokuma Wachi Shintaro Wakasugi Tomoya Wakayama Haonan Wang Jing Wang 0023 Leyang Wang Liwei Wang 0001 Mingze Wang Yakun Wang 0005 Zhangyang Wang Zhichao Wang Chihiro Watanabe Kishan Wimalawarne Hau-San Wong Denny Wu Chenyuan Xu Ji Xu Zhiqiang Xu 0003 Bo Xue 0004 Atsushi Yaguchi Makoto Yamada Shoichiro Yamaguchi Kakei Yamamoto Masao Yamanaka Kenji Yamanishi Atsushi Yamashita Junchi Yan Hiroki Yanagisawa Greg Yang Zhuoran Yang Shingo Yashima Rio Yokota Kazuo Yonekura Masatoshi Yukishima Ce Zhang 0001 Miao Zhang 0022 Qingfu Zhang 0001 Qixin Zhang Tianzong Zhang Tongcheng Zhang Yuanhe Zhang Huanjian Zhou Zhanpeng Zhou Jun Zhu 0001 Difan Zou