Mingyi Hong
Papers
2
Total Citations
18
H-Index
2
About
Mingyi Hong is a prominent researcher whose work spans the intersections of multi-agent reinforcement learning, optimization theory, and distributed systems. His research addresses some of the most challenging problems in modern machine learning and control, with a particular focus on developing mathematically rigorous algorithms for complex, real-world applications. Among his notable contributions, Hong has advanced the field of multi-agent reinforcement learning (MARL), tackling the inherent difficulties of coordinating multiple interacting agents in decentralized settings such as sensor networks, swarm robotics, and power grids. His 2018 work on policy evaluation via double averaging primal-dual optimization demonstrated innovative approaches to stabilizing learning in these complex environments, earning 14 citations. More recently, his 2022 research on Maximum-Likelihood Inverse Reinforcement Learning introduced finite-time guarantees to IRL algorithms, addressing the computationally expensive nested optimization structures that have long challenged the field. Hong's work is distinguished by its emphasis on theoretical rigor — providing convergence guarantees and finite-time analyses that bridge the gap between practical algorithm design and mathematical foundations. His contributions are particularly valuable for students and researchers seeking principled approaches to reinforcement learning in distributed and multi-agent environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2