Papers

Reinforcement learning

  • Miyaguchi, K. (2024). Worst-Case Offline Reinforcement Learning with Arbitrary Data Support. NeurIPS-24. [poster/paper]
  • Kajino, H., Miyaguchi, K., & Osogami, T. (2023). Biases in evaluation of molecular optimization methods and bias reduction strategies. ICML-23.
  • Miyaguchi, K. (2022). A Theoretical Framework of Almost Hyperparameter-free Hyperparameter Selection Methods for Offline Policy Evaluation. AAAI-22 Workshop. [paper]
  • Miyaguchi, K. (2021). Asymptotically Exact Error Characterization of Offline Policy Evaluation with Linear Direct Method under Unrealizability. NeurIPS-21. [poster/paper]

Machine learning theory

  • Kobayashi, M., Miyaguchi, K., & Matsushima, S. (2022). Detection of Unobserved Common Cause in Discrete Data Based on the MDL Principle. BigData-22.
  • Miyaguchi, K., Katsuki, T., Koseki, A., & Iwamori, T. (2022). Variational Inference for Discriminative Learning with Generative Modeling of Feature Incompletion. ICLR-22 (Oral). [poster/paper]
  • Miyaguchi, K. (2019). PAC-Bayesian transportation bound. arXiv preprint. [paper]
  • Miyaguchi, K. & Yamanishi, K. (2019). Adaptive Minimax Regret against Smooth Logarithmic Losses over High-Dimensional l1-Balls via Envelope Complexity. AISTATS-19. [paper]
  • Miyaguchi, K., & Yamanishi, K. (2018). High-dimensional penalty selection via minimum description length principle. Machine Learning (journal). [paper]
  • Miyaguchi, K., Matsushima, S., & Yamanishi, K. (2017). Sparse Graphical Modeling via Stochastic Complexity. SDM-17. [paper]
  • Suzuki, A., Miyaguchi, K., & Yamanishi, K. (2016). Structure selection for convolutive non-negative matrix factorization using normalized maximum likelihood coding. ICDM-16.

Time series

  • Katsuki, T., Miyaguchi, K., Koseki, A., Iwamori, T., Yanagiya, R., & Suzuki, A. (2022). Cumulative Stay-time Representation for Electronic Health Records in Medical Event Time Prediction. IJCAI-22.
  • Miyaguchi, K. & Kajino, K. (2019). Cogra: Concept-drift-aware Stochastic Gradient Descent for Time-series Forecasting. AAAI-19. [paper]
  • Kaneko, R., Miyaguchi, K., & Yamanishi, K. (2017). Detecting changes in streaming data with information-theoretic windowing. BigData-17. [paper]
  • Yamanishi, K., & Miyaguchi, K. (2016). Detecting gradual changes from data stream using MDL-change statistics. BigData-16.
  • Miyaguchi, K., & Yamanishi, K. (2015). On-line detection of continuous changes in stochastic processes. DSAA-15. [paper]

Game theory

  • Kinoshita, H., Osogami, T., & Miyaguchi, K. (2024). Socially efficient mechanism on the minimum budget. arXiv.

Invited Talks

  • Miyaguchi, K. (2024). 限られたデータを用いた強化学習の成功条件について. 2024年度CIGS経済・社会の分野横断的研究会. [slide]
  • Miyaguchi, K. (2022). 汎化誤差解析から始める統計的学習理論 (Tutorial on Statistical Learning Theory with Generalization Error Analysis). IBIS Workshop 2022 tutorial. [slide]
  • Miyaguchi, K. (2022). NeurIPS2021 Review on Deep Learning Theories, The 83rd JSAI Seminar.
  • Miyaguchi, K. (2020). PAC-Bayesian Transportation Bound, RIKEN AIP Mathematical Seminar.

Books

  • Kajino, H., Miyaguchi, K., Osogami, T., Iwaki, R., and Wachi, A. (2024). 強化学習から信頼できる意思決定へ (From reinforcement learning to reliable decision making), Yamanishi, K. Ed., Saiensu-sha. [publisher/amazon]