Yoshimasa Tsuruoka

About

Yoshimasa Tsuruoka is a leading researcher in reinforcement learning (RL) and control systems, with a focus on making AI-driven automation scalable, robust, and practical for real-world industrial applications. His work bridges the gap between theoretical RL advances and the demanding requirements of large-scale process control. In his highly influential paper, *“Local Control is All You Need: Decentralizing and Coordinating Reinforcement Learning for Large-Scale Process Control”* (2022, 4 citations), Tsuruoka challenges the centralized paradigm by proposing a decentralized RL framework that coordinates multiple local agents, dramatically improving scalability and generalization in complex industrial environments. This work has been recognized for its potential to reduce the need for constant parameter tuning, a long-standing pain point in process industries. Tsuruoka also tackles the critical issue of robustness in hierarchical RL. In *“Learning Robust Options by Conditional Value at Risk Optimization”* (2019, 2 citations), he introduces a novel risk-aware method for learning options—temporally extended actions—that are resilient to uncertain model parameters. By optimizing for Conditional Value at Risk (CVaR), his approach moves beyond worst-case or average-case thinking, offering a principled way to balance performance and safety. This contribution is particularly valuable for high-stakes domains like chemical processing and robotics, where model inaccuracies can lead to catastrophic failures. Tsuruoka’s work is shaping the next generation of RL systems that are both intelligent and dependable.

Research Focus

Key Achievements

Top Papers

1. 1
   Local Control is All You Need: Decentralizing and Coordinating Reinforcement Learning for Large-Scale Process Control

   4 citations · 2022
2. 2
   Learning Robust Options by Conditional Value at Risk Optimization

   2 citations · 2019

Key Collaborators

Contact & Links