Shinsaku Sakaue
Shinsaku Sakaue

About Me

I am a research scientist at CyberAgent AI Lab. Previously, I was a visiting scientist at RIKEN AIP, project assistant professor at UTokyo IST, and researcher at NTT CS Lab.

Interests
  • Discrete/Continuous Optimization
  • Online Learning
  • Algorithmic Game Theory
Education

  • PhD Informatics

    Kyoto University

  • MSc Information Science and Technology

    The University of Tokyo

  • BEng Mathematical Engineering

    The University of Tokyo

Experience

  1. Research Scientist

    CyberAgent AI Lab

  2. Visiting Scientist

    RIKEN Center for Advanced Intelligence Project

  3. Project Assistant Professor

    The University of Tokyo

  4. Researcher

    NTT Communication Science Laboratories

Education

  1. PhD Informatics

    Kyoto University

  2. MSc Information Science and Technology

    The University of Tokyo

  3. BEng Mathematical Engineering

    The University of Tokyo
Awards
IEICE TC-IBISML Research Award 2017
IBISML ∙ November 2018
Teaching
  • Exercise course of geometry (2020–2024 Autumn, The University of Tokyo)
  • Short exercise course of numerical methods (2020 Spring & Autumn, The University of Tokyo)
  • Short exercise course of discrete methods (2022–2024 Spring & Autumn, The University of Tokyo)
Featured Papers
Online structured prediction with Fenchel–Young losses and improved surrogate regret for online multiclass classification with logistic loss
Discrete-convex-analysis-based framework for warm-starting algorithms with predictions
All Papers
Tomoharu Iwata, Shinsaku Sakaue (2025). Learning to generate projections for reducing dimensionality of heterogeneous linear programming problems. International Conference on Machine Learning (ICML), to appear.
Han Bao, Shinsaku Sakaue (2025). Inverse optimization with prediction market: A characterization of scoring rules for elciting system states. International Conference on Artificial Intelligence and Statistics (AISTATS).
PDF
Shinsaku Sakaue, Han Bao, Taira Tsuchiya (2025). Revisiting online learning approach to inverse linear optimization: A Fenchel–Young loss perspective and gap-dependent regret analysis. International Conference on Artificial Intelligence and Statistics (AISTATS).
PDF
Yuki Shibukawa, Taira Tsuchiya, Shinsaku Sakaue, Kenji Yamanishi (2025). Bandit and delayed feedback in online structured prediction. arXiv [cs.LG, stat.ML].
PDF
Han Bao, Shinsaku Sakaue, Yuki Takezawa (2025). Any-stepsize gradient descent for separable data under Fenchel–Young losses. arXiv [stat.ML, cs.LG].
PDF
See all publications