Masashi Sugiyama - researchr alias

researchr

You are not signed in
Sign in
Sign up

Viewing Publication 1 - 100 from 441

2023

Is the Performance of My Deep Network Too Good to Be True? A Direct Approach to Estimating the Bayes Error in Binary ClassificationTakashi Ishida 0001, Ikko Yamane, Nontawat Charoenphakdee, Gang Niu 0001, Masashi Sugiyama. iclr 2023: [doi]

Learning Intention-Aware Policies in Deep Reinforcement LearningTingting Zhao 0001, S. Wu, G. Li, Y. Chen, Gang Niu 0001, Masashi Sugiyama. neco, 35(10):1657-1677, October 2023. [doi]

Representation learning for continuous action spaces is beneficial for efficient policy learningTingting Zhao 0001, Ying Wang, Wei Sun, Yarui Chen, Gang Niu 0001, Masashi Sugiyama. NN, 159:137-152, February 2023. [doi]

Class-Wise Denoising for Robust Learning Under Label NoiseChen Gong 0002, Yongliang Ding, Bo Han 0003, Gang Niu 0001, Jian Yang 0003, Jane You, Dacheng Tao, Masashi Sugiyama. pami, 45(3):2835-2848, March 2023. [doi]

Seeing Differently, Acting Similarly: Heterogeneously Observable Imitation LearningXin-Qiang Cai, Yao-Xiang Ding, Zi-Xuan Chen, Yuan Jiang 0001, Masashi Sugiyama, Zhi-Hua Zhou. iclr 2023: [doi]

Diversity-enhancing Generative Network for Few-shot Hypothesis AdaptationRuijiang Dong, Feng Liu, Haoang Chi, Tongliang Liu, Mingming Gong, Gang Niu 0001, Masashi Sugiyama, Bo Han 0003. icml 2023: 8260-8275 [doi]

GAT: Guided Adversarial Training with Pareto-optimal Auxiliary TasksSalah Ghamizi, Jingfeng Zhang, Maxime Cordy, Mike Papadakis, Masashi Sugiyama, Yves Le Traon. icml 2023: 11255-11282 [doi]

Audio Signal Enhancement with Learning from Positive and Unlabeled DataNobutaka Ito, Masashi Sugiyama. icassp 2023: 1-5 [doi]

Optimality of Thompson Sampling with Noninformative Priors for Pareto BanditsJongyeong Lee, Junya Honda, Chao-Kai Chiang, Masashi Sugiyama. icml 2023: 18810-18851 [doi]

Positive-unlabeled classification under class-prior shift: a prior-invariant approach based on density ratio estimationShota Nakajima, Masashi Sugiyama. ml, 112(3):889-919, March 2023. [doi]

Root cause estimation of faults in production processes: a novel approach inspired by approximate Bayesian computationYosuke Otsubo, Naoya Otani, Megumi Chikasue, Mineyuki Nishino, Masashi Sugiyama. ijpr, 61(5):1556-1574, March 2023. [doi]

Learning With Proper Partial LabelsZhenguo Wu, Jiaqi Lv, Masashi Sugiyama. neco, 35(1):58-81, January 2023. [doi]

A Category-theoretical Meta-analysis of Definitions of DisentanglementYivan Zhang, Masashi Sugiyama. icml 2023: 41596-41612 [doi]

2022

Pairwise Supervision Can Provably Elicit a Decision BoundaryHan Bao 0002, Takuya Shimada, Liyuan Xu, Issei Sato, Masashi Sugiyama. aistats 2022: 2618-2640 [doi]

Adversarial Training with Complementary Labels: On the Benefit of Gradually Informative AttacksJianan Zhou 0002, Jianing Zhu, Jingfeng Zhang, Tongliang Liu, Gang Niu 0001, Bo Han 0003, Masashi Sugiyama. nips 2022: [doi]

Centroid Estimation With Guaranteed Efficiency: A General Framework for Weakly Supervised LearningChen Gong 0002, Jian Yang 0003, Jane You, Masashi Sugiyama. pami, 44(6):2841-2855, 2022. [doi]

Learning Contrastive Embedding in Low-Dimensional SpaceShuo Chen 0003, Chen Gong 0002, Jun Li 0027, Jian Yang 0003, Gang Niu 0001, Masashi Sugiyama. nips 2022: [doi]

Adapting to Online Label Shift with Provable GuaranteesYong Bai, Yu-Jie Zhang, Peng Zhao 0006, Masashi Sugiyama, Zhi-Hua Zhou. nips 2022: [doi]

Generalizing Consistent Multi-Class Classification with Rejection to be Compatible with Arbitrary LossesYuzhou Cao, Tianchi Cai, Lei Feng 0006, Lihong Gu, Jinjie Gu, Bo An 0001, Gang Niu 0001, Masashi Sugiyama. nips 2022: [doi]

Instance-Dependent Label-Noise Learning with Manifold-Regularized Transition Matrix EstimationDe Cheng, Tongliang Liu, Yixiong Ning, Nannan Wang 0001, Bo Han 0003, Gang Niu 0001, Xinbo Gao 0001, Masashi Sugiyama. cvpr 2022: 16609-16618 [doi]

Meta Discovery: Learning to Discover Novel Classes given Very Limited DataHaoang Chi, Feng Liu, Wenjing Yang, Long Lan, Tongliang Liu, Bo Han 0003, Gang Niu 0001, Mingyuan Zhou, Masashi Sugiyama. iclr 2022: [doi]

Synergy-of-Experts: Collaborate to Improve Adversarial RobustnessSen Cui, Jingfeng Zhang, Jian Liang, Bo Han 0003, Masashi Sugiyama, Changshui Zhang. nips 2022: [doi]

Neural Networks special issue on Artificial Intelligence and Brain ScienceKenji Doya, Karl J. Friston, Masashi Sugiyama, Joshua B. Tenenbaum. NN, 155:328-329, 2022. [doi]

Predictive variational Bayesian inference as risk-seeking optimizationFutoshi Futami, Tomoharu Iwata, Naonori Ueda, Issei Sato, Masashi Sugiyama. aistats 2022: 5051-5083 [doi]

Learning from Noisy Complementary Labels with Robust Loss FunctionsHiroki Ishiguro, Takashi Ishida 0001, Masashi Sugiyama. ieicetd, 105(2):364-376, 2022. [doi]

LocalDrop: A Hybrid Regularization for Deep Neural NetworksZiqing Lu, Chang Xu 0002, Bo Du 0001, Takashi Ishida 0001, Lefei Zhang, Masashi Sugiyama. pami, 44(7):3590-3601, 2022. [doi]

Federated Learning from Only Unlabeled Data with Class-conditional-sharing ClientsNan Lu, Zhao Wang, Xiaoxiao Li, Gang Niu 0001, Qi Dou, Masashi Sugiyama. iclr 2022: [doi]

Deep learning, reinforcement learning, and world modelsYutaka Matsuo, Yann LeCun, Maneesh Sahani, Doina Precup, David Silver, Masashi Sugiyama, Eiji Uchibe, Jun Morimoto. NN, 152:267-275, 2022. [doi]

Robust computation of optimal transport by β-potential regularizationShintaro Nakamura, Han Bao 0002, Masashi Sugiyama. acml 2022: 770-785 [doi]

Discovering diverse solutions in deep reinforcement learning by maximizing state-action-based mutual informationTakayuki Osa, Voot Tangkaratt, Masashi Sugiyama. NN, 152:90-104, 2022. [doi]

Fast and Robust Rank Aggregation against Model MisspecificationYuangang Pan, Ivor W. Tsang, Weijie Chen 0001, Gang Niu 0001, Masashi Sugiyama. jmlr, 23, 2022. [doi]

Learning and Mining with Noisy LabelsMasashi Sugiyama, Tongliang Liu, Bo Han 0003, Yang Liu, Gang Niu 0001. CIKM 2022: 5152-5155 [doi]

Multi-class Classification from Multiple Unlabeled Datasets with Partial Risk RegularizationYuting Tang, Nan Lu, Tianyi Zhang, Masashi Sugiyama. acml 2022: 990-1005 [doi]

Improving imbalanced classification using near-miss instancesAkira Tanimoto, So Yamada, Takashi Takenouchi, Masashi Sugiyama, Hisashi Kashima. eswa, 201:117130, 2022. [doi]

To Smooth or Not? When Label Smoothing Meets Noisy LabelsJiaheng Wei, Hangyu Liu, Tongliang Liu, Gang Niu 0001, Masashi Sugiyama, Yang Liu. icml 2022: 23589-23614 [doi]

Learning from Noisy Pairwise Similarity and Unlabeled DataSonghua Wu, Tongliang Liu, Bo Han 0003, Jun Yu 0001, Gang Niu 0001, Masashi Sugiyama. jmlr, 23, 2022. [doi]

Sample Selection with Uncertainty of Losses for Learning with Noisy LabelsXiaobo Xia, Tongliang Liu, Bo Han 0003, Mingming Gong, Jun Yu, Gang Niu 0001, Masashi Sugiyama. iclr 2022: [doi]

Adaptive Inertia: Disentangling the Effects of Adaptive Learning Rate and MomentumZeke Xie, Xinrui Wang, Huishuai Zhang, Issei Sato, Masashi Sugiyama. icml 2022: 24430-24459 [doi]

Adversarial Attack and Defense for Non-Parametric Two-Sample TestsXilie Xu, Jingfeng Zhang, Feng Liu, Masashi Sugiyama, Mohan S. Kankanhalli. icml 2022: 24743-24769 [doi]

Towards Adversarially Robust Deep Image DenoisingHanshu Yan, Jingfeng Zhang, Jiashi Feng, Masashi Sugiyama, Vincent Y. F. Tan. IJCAI 2022: 1516-1522 [doi]

Rethinking Class-Prior Estimation for Positive-Unlabeled LearningYu Yao, Tongliang Liu, Bo Han 0003, Mingming Gong, Gang Niu 0001, Masashi Sugiyama, Dacheng Tao. iclr 2022: [doi]

NoiLin: Improving adversarial training and correcting stereotype of noisy labelsJingfeng Zhang, Xilie Xu, Bo Han 0003, Tongliang Liu, LiZhen Cui, Gang Niu 0001, Masashi Sugiyama. tmlr, 2022, 2022. [doi]

Exploiting Class Activation Value for Partial-Label LearningFei Zhang, Lei Feng, Bo Han 0003, Tongliang Liu, Gang Niu 0001, Tao Qin, Masashi Sugiyama. iclr 2022: [doi]

2021

Fenchel-Young Losses with Skewed Entropies for Class-posterior Probability EstimationHan Bao 0002, Masashi Sugiyama. aistats 2021: 1648-1656 [doi]

Large-Margin Contrastive Learning with Distance Polarization RegularizerShuo Chen 0003, Gang Niu 0001, Chen Gong 0002, Jun Li, Jian Yang, Masashi Sugiyama. icml 2021: 1673-1683 [doi]

Confidence Scores Make Instance-dependent Label-noise Learning PossibleAntonin Berthon, Bo Han 0003, Gang Niu 0001, Tongliang Liu, Masashi Sugiyama. icml 2021: 825-836 [doi]

Learning from Similarity-Confidence DataYuzhou Cao, Lei Feng, Yitian Xu, Bo An 0001, Gang Niu 0001, Masashi Sugiyama. icml 2021: 1272-1282 [doi]

Classification with Rejection Based on Cost-sensitive ClassificationNontawat Charoenphakdee, Zhenghang Cui, Yivan Zhang, Masashi Sugiyama. icml 2021: 1507-1517 [doi]

On Focal Loss for Class-Posterior Probability Estimation: A Theoretical PerspectiveNontawat Charoenphakdee, Jayakorn Vongkulbhisal, Nuttapong Chairatanakul, Masashi Sugiyama. cvpr 2021: 5202-5211 [doi]

Learning from Noisy Similar and Dissimilar DataSoham Dan, Han Bao 0002, Masashi Sugiyama. pkdd 2021: 233-249 [doi]

Learning Diverse-Structured Networks for Adversarial RobustnessXuefeng Du, Jingfeng Zhang, Bo Han 0003, Tongliang Liu, Yu Rong, Gang Niu 0001, JunZhou Huang, Masashi Sugiyama. icml 2021: 2880-2891 [doi]

Pointwise Binary Classification with Pairwise Confidence ComparisonsLei Feng, Senlin Shu, Nan Lu, Bo Han 0003, Miao Xu, Gang Niu 0001, Bo An 0001, Masashi Sugiyama. icml 2021: 3252-3262 [doi]

γ-ABC: Outlier-Robust Approximate Bayesian Computation Based on a Robust Divergence EstimatorMasahiro Fujisawa, Takeshi Teshima, Issei Sato, Masashi Sugiyama. aistats 2021: 1783-1791 [doi]

Loss function based second-order Jensen inequality and its application to particle variational inferenceFutoshi Futami, Tomoharu Iwata, Naonori Ueda, Issei Sato, Masashi Sugiyama. nips 2021: 6803-6815 [doi]

Maximum Mean Discrepancy Test is Aware of Adversarial AttacksRuize Gao, Feng Liu, Jingfeng Zhang, Bo Han 0003, Tongliang Liu, Gang Niu 0001, Masashi Sugiyama. icml 2021: 3564-3575 [doi]

Scalable Evaluation and Improvement of Document Set Expansion via Neural Positive-Unlabeled LearningAlon Jacovi, Gang Niu 0001, Yoav Goldberg, Masashi Sugiyama. eacl 2021: 581-592 [doi]

Provably End-to-end Label-noise Learning without Anchor PointsXuefeng Li, Tongliang Liu, Bo Han 0003, Gang Niu 0001, Masashi Sugiyama. icml 2021: 6403-6413 [doi]

Binary Classification from Multiple Unlabeled Datasets via Surrogate Set ClassificationNan Lu, Shida Lei, Gang Niu 0001, Issei Sato, Masashi Sugiyama. icml 2021: 7134-7144 [doi]

Constraint learning for control tasks with limited duration barrier functionsMotoya Ohnishi, Gennaro Notomista, Masashi Sugiyama, Magnus Egerstedt. automatica, 127:109504, 2021. [doi]

A unified view of likelihood ratio and reparameterization gradientsPaavo Parmas, Masashi Sugiyama. aistats 2021: 4078-4086 [doi]

Information-Theoretic Representation Learning for Positive-Unlabeled ClassificationTomoya Sakai, Gang Niu 0001, Masashi Sugiyama. neco, 33(1):244-268, 2021. [doi]

Classification From Pairwise Similarities/Dissimilarities and Unlabeled Data via Empirical Risk MinimizationTakuya Shimada, Han Bao 0002, Issei Sato, Masashi Sugiyama. neco, 33(5):1234-1268, 2021. [doi]

Mixture Proportion Estimation in Weakly Supervised LearningMasashi Sugiyama. CIKM 2021: [doi]

Robust Imitation Learning from Noisy DemonstrationsVoot Tangkaratt, Nontawat Charoenphakdee, Masashi Sugiyama. aistats 2021: 298-306 [doi]

Incorporating causal graphical prior knowledge into predictive modeling via simple data augmentationTakeshi Teshima, Masashi Sugiyama. uai 2021: 86-96 [doi]

Semisupervised Ordinal Regression Based on Empirical Risk MinimizationTaira Tsuchiya, Nontawat Charoenphakdee, Issei Sato, Masashi Sugiyama. neco, 33(12):3361-3412, 2021. [doi]

Probabilistic Margins for Instance Reweighting in Adversarial TrainingQizhou Wang, Feng Liu 0003, Bo Han 0003, Tongliang Liu, Chen Gong 0002, Gang Niu 0001, Mingyuan Zhou, Masashi Sugiyama. nips 2021: 23258-23269 [doi]

A Diffusion Theory For Deep Learning Dynamics: Stochastic Gradient Descent Exponentially Favors Flat MinimaZeke Xie, Issei Sato, Masashi Sugiyama. iclr 2021: [doi]

Artificial Neural Variability for Deep Learning: On Overfitting, Noise Memorization, and Catastrophic ForgettingZeke Xie, Fengxiang He, Shaopeng Fu, Issei Sato, Dacheng Tao, Masashi Sugiyama. neco, 33(8):2163-2192, 2021. [doi]

Positive-Negative Momentum: Manipulating Stochastic Gradient Noise to Improve GeneralizationZeke Xie, Li Yuan, Zhanxing Zhu, Masashi Sugiyama. icml 2021: 11448-11458 [doi]

Direction Matters: On Influence-Preserving Graph Summarization and Max-Cut Principle for Directed GraphsWenkai Xu, Gang Niu 0001, Aapo Hyvärinen, Masashi Sugiyama. neco, 33(8):2128-2162, 2021. [doi]

Mediated Uncoupled Learning: Learning Functions without Direct Input-output CorrespondencesIkko Yamane, Junya Honda, Florian Yger, Masashi Sugiyama. icml 2021: 11637-11647 [doi]

CIFS: Improving Adversarial Robustness of CNNs via Channel-wise Importance-based Feature SelectionHanshu Yan, Jingfeng Zhang, Gang Niu 0001, Jiashi Feng, Vincent Y. F. Tan, Masashi Sugiyama. icml 2021: 11693-11703 [doi]

Lower-Bounded Proper Losses for Weakly Supervised ClassificationShuhei M. Yoshida, Takashi Takenouchi, Masashi Sugiyama. icml 2021: 12110-12120 [doi]

Learning Noise Transition Matrix from Only Noisy Labels via Total Variation RegularizationYivan Zhang, Gang Niu 0001, Masashi Sugiyama. icml 2021: 12501-12512 [doi]

Geometry-aware Instance-reweighted Adversarial TrainingJingfeng Zhang, Jianing Zhu, Gang Niu 0001, Bo Han 0003, Masashi Sugiyama, Mohan S. Kankanhalli. iclr 2021: [doi]

A One-Step Approach to Covariate Shift AdaptationTianyi Zhang, Ikko Yamane, Nan Lu, Masashi Sugiyama. sncs, 2(4):319, 2021. [doi]

2020

Calibrated Surrogate Losses for Adversarially Robust ClassificationHan Bao, Clayton Scott, Masashi Sugiyama. colt 2020: 408-451 [doi]

Calibrated Surrogate Maximization of Linear-fractional Utility in Binary ClassificationHan Bao, Masashi Sugiyama. aistats 2020: 2337-2347 [doi]

Active deep Q-learning with demonstrationSi-An Chen, Voot Tangkaratt, Hsuan-Tien Lin, Masashi Sugiyama. ml, 109(9):1699-1725, 2020. [doi]

Unbiased Risk Estimators Can Mislead: A Case Study of Learning with Complementary LabelsYu-Ting Chou, Gang Niu 0001, Hsuan-Tien Lin, Masashi Sugiyama. icml 2020: 1929-1938 [doi]

Classification from Triplet Comparison DataZhenghang Cui, Nontawat Charoenphakdee, Issei Sato, Masashi Sugiyama. neco, 32(3):659-681, 2020. [doi]

Learning with Multiple Complementary LabelsLei Feng, Takuo Kaneko, Bo Han 0003, Gang Niu 0001, Bo An 0001, Masashi Sugiyama. icml 2020: 3072-3081 [doi]

Accelerating the diffusion-based ensemble sampling by non-reversible dynamicsFutoshi Futami, Issei Sato, Masashi Sugiyama. icml 2020: 3337-3347 [doi]

Do We Need Zero Training Loss After Achieving Zero Training Error?Takashi Ishida, Ikko Yamane, Tomoya Sakai, Gang Niu 0001, Masashi Sugiyama. icml 2020: 4604-4614 [doi]

Partially Zero-shot Domain Adaptation from Incomplete Target Data with Missing ClassesMasato Ishii, Takashi Takenouchi, Masashi Sugiyama. wacv 2020: 3041-3049 [doi]

Online Dense Subgraph Discovery via Blurred-Graph FeedbackYuko Kuroki, Atsushi Miyauchi 0001, Junya Honda, Masashi Sugiyama. icml 2020: 5522-5532 [doi]

Polynomial-Time Algorithms for Multiple-Arm Identification with Full-Bandit FeedbackYuko Kuroki, Liyuan Xu, Atsushi Miyauchi 0001, Junya Honda, Masashi Sugiyama. neco, 32(9):1733-1773, 2020. [doi]

Principled analytic classifier for positive-unlabeled learning via weighted integral probability metricYongchan Kwon, Wonyoung Kim, Masashi Sugiyama, Myunghee Cho Paik. ml, 109(3):513-532, 2020. [doi]

Are Registration Uncertainty and Error Monotonically Associated?Jie Luo, Sarah F. Frisken, Duo Wang, Alexandra J. Golby, Masashi Sugiyama, William M. Wells III. miccai 2020: 264-274 [doi]

Calibrated Surrogate Maximization of DiceMarcus Nordström, Han Bao, Fredrik Löfman, Henrik Hult, Atsuto Maki, Masashi Sugiyama. miccai 2020: 269-278 [doi]

Binary classification with ambiguous training dataNaoya Otani, Yosuke Otsubo, Tetsuya Koike, Masashi Sugiyama. ml, 109(12):2369-2388, 2020. [doi]

Unsupervised key frame selection using information theory and colour histogram differenceJanya Sainui, Masashi Sugiyama. ijbidm, 16(3):324-344, 2020. [doi]

Binary Classification from Positive Data with Skewed ConfidenceKazuhiko Shinoda, Hirotaka Kaji, Masashi Sugiyama. IJCAI 2020: 3328-3334 [doi]

Simultaneous Planning for Item Picking and Placing by Deep Reinforcement LearningTatsuya Tanaka, Toshimitsu Kaneko, Masahiro Sekine, Voot Tangkaratt, Masashi Sugiyama. iros 2020: 9705-9711 [doi]

Few-shot Domain Adaptation by Causal Mechanism TransferTakeshi Teshima, Issei Sato, Masashi Sugiyama. icml 2020: 9458-9469 [doi]

Coupling-based Invertible Neural Networks Are Universal Diffeomorphism ApproximatorsTakeshi Teshima, Isao Ishikawa, Koichi Tojo, Kenta Oono, Masahiro Ikeda, Masashi Sugiyama. nips 2020: [doi]

Analysis and Design of Thompson Sampling for Stochastic Partial MonitoringTaira Tsuchiya, Junya Honda, Masashi Sugiyama. nips 2020: [doi]

Normalized Flat Minima: Exploring Scale Invariant Definition of Flat Minima for Neural Networks Using {PACYusuke Tsuzuku, Issei Sato, Masashi Sugiyama. icml 2020: 9636-9647 [doi]

Learning from Aggregate ObservationsYivan Zhang, Nontawat Charoenphakdee, Zhenguo Wu, Masashi Sugiyama. nips 2020: [doi]

All Publications

Filter by Year

[-]
OR AND NOT 1

Filter by Type
OR AND NOT 1

Filter by Tag

[+]
OR AND NOT 1

Filter by Venue

[+]
OR AND NOT 1

Filter by Author

[+]
OR AND NOT 1

Filter by Top Terms

[+]
OR AND NOT 1

Researchr

Publications by 'Masashi Sugiyama'

Viewing Publication 1 - 100 from 441

2023

2022

2021

2020

All Publications

Filter by Year [-]OR AND NOT 1

Filter by TypeOR AND NOT 1

Filter by Tag [+]OR AND NOT 1

Filter by Venue [+]OR AND NOT 1

Filter by Author [+]OR AND NOT 1

Filter by Top Terms [+]OR AND NOT 1

Researchr

Publications by 'Masashi Sugiyama'

Viewing Publication 1 - 100 from 441

2023

2022

2021

2020

Filter by Year

[-]
OR AND NOT 1

Filter by Type
OR AND NOT 1

Filter by Tag

[+]
OR AND NOT 1

Filter by Venue

[+]
OR AND NOT 1

Filter by Author

[+]
OR AND NOT 1

Filter by Top Terms

[+]
OR AND NOT 1