革新知能統合研究センター

学習理論に関する国際会議 第38回 Annual Conference on Learning Theory (COLT 2025) においてAIPセンターから論文が4本採択されました。

[Website] https://learningtheory.org/colt2025/

• Corrupted Learning Dynamics in Games.
  Taira Tsuchiya (The University of Tokyo / RIKEN AIP)
  Shinji Ito (The University of Tokyo / RIKEN AIP)
  Haipeng Luo (University of Southern California)

• Data-dependent Bounds with T-Optimal Best-of-Both-Worlds Guarantees in Multi-Armed Bandits using Stability-Penalty Matching.
  Quan Nguyen* (University of Victoria)
  Shinji Ito (The University of Tokyo / RIKEN AIP)
  Junpei Komiyama (New York University)
  Nishant A. Mehta (University of Victoria)
  * The work was done when he was an intern at RIKEN AIP.
• Instance-Dependent Regret Bounds for Learning Two-Player Zero-Sum Games with Bandit Feedback.
  Shinji Ito (The University of Tokyo / RIKEN AIP)
  Haipeng Luo (University of Southern California)
  Taira Tsuchiya (The University of Tokyo / RIKEN AIP)
  Yue Wu (University of Southern California)
• The Adaptive Complexity of Finding a Stationary Point
  Huanjian Zhou (The University of Tokyo / RIKEN AIP)
  Andi Han (RIKEN AIP)
  Akiko Takeda (The University of Tokyo / RIKEN AIP)
  Masashi Sugiyama (RIKEN AIP / The University of Tokyo)

関連研究室