1. Home
  2. Research
  3. Centers & Labs
  4. RIKEN Center for Advanced Intelligence Project
  5. Generic Technology Research Group

RIKEN Center for Advanced Intelligence Project Imperfect Information Learning Team

Team Leader: Masashi Sugiyama (D.Eng.)

Research Summary

Masashi  Sugiyama(D.Eng.)

Recently, machine learning technology with big data has been actively investigated and its effectiveness has been demonstrated. However, depending on application domains, it is difficult or even it is not possible to collect a large amount of data. In the Imperfect Information Learning Team, for various machine learning tasks including supervised learning, unsupervised learning, and reinforcement learning, we develop novel algorithms that allow accurate learning from limited information. We also elucidate their theoretical properties and apply them to various real-world applications ranging from fundamental science to business.

Research Subjects:

  • Development of machine learning algorithms from imperfect information
  • Theoretical analysis of machine learning algorithms
  • Real-world application of machine learning algorithms

Main Research Fields

  • Informatics

Related Research Fields

  • Intelligent informatics
  • Perceptual information processing
  • Statistical science

Keywords

  • artificial intelligence
  • machine learning
  • weakly supervised learning
  • reinforcement learning
  • deep learning

Selected Publications

  • 1. Zhang, J., Xu, X., Han, B., Niu, G., Cui, L., Sugiyama, M., & Kankanhalli, M.:
    "Attacks which do not kill training make adversarial learning stronger."
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 2. Tangkaratt, V., Han, B., Khan, M. E., & Sugiyama, M.:
    "Variational imitation learning with diverse-quality demonstrations."
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 3. Han, B., Niu, G., Yu, X., Yao, Q., Xu, M., Tsang, I., & Sugiyama, M.:
    "SIGUA: Forgetting may make learning with noisy labels more robust."
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 4. Feng, L., Kaneko, T., Han, B., Niu, G., An, B., & Sugiyama, M.:
    "Learning with multiple complementary labels."
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 5. Chou, Y.-T., Niu, G., Lin, H.-T., & Sugiyama, M.:
    "Unbiased risk estimators can mislead: A case study of learning with complementary labels."
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 6. Lv, J., Xu, M., Feng, L., Niu, G., Geng, X., & Sugiyama, M.:
    "Progressive identification of true labels for partial-label learning."
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 7. Ishida, T., Yamane, I., Sakai, T., Niu, G., & Sugiyama, M.:
    "Do we need zero training loss after achieving zero training error?"
    In Proceedings of 37th International Conference on Machine Learning (ICML2020), to appear.
  • 8. Lu, N., Zhang, T., Niu, G., & Sugiyama, M.:
    Mitigating overfitting in supervised classification from two "unlabeled datasets: A consistent risk correction approach."
    In Proceedings of 23rd International Conference on Artificial Intelligence and Statistics (AISTATS2020), pp.1115-1125, 2020.
  • 9. Xia, X., Liu, T., Wang, N., Han, B., Gong, C., Niu, G., & Sugiyama, M.:
    "Are anchor points really indispensable in label-noise learning?"
    In Advances in Neural Information Processing Systems 32 (NeurIPS2019), pp.6835-6846, 2019.
  • 10. Wu, Y.-H., Charoenphakdee, N., Bao, H., Tangkaratt, V., & Sugiyama, M.:
    "Imitation learning from imperfect demonstration."
    In Proceedings of 36th International Conference on Machine Learning (ICML2019), pp.6818-6827, 2019.

Recent Research Results

Related Links

Lab Members

Principal investigator

Masashi Sugiyama
Team Leader

Core members

Gang Niu
Research Scientist
Voot Tangkaratt
Postdoctoral Researcher

Contact Information

Nihonbashi 1-chome Mitsui Building, 15th floor,
1-4-1 Nihonbashi,
Chuo-ku, Tokyo
103-0027, Japan
Email: masashi.sugiyama [at] riken.jp

Top