List of English Publication

@inproceedings{nokey,

title = {Distributed Q-Learning on Multi-Agent Markov Decision Process with Heterogeneous State Transition Probabilities},

author = {Yong Joo Do and Deuksun Hong and Hyungbo Shim},

url = {https://ieeexplore.ieee.org/document/11312118},

doi = {10.1109/CDC57313.2025.11312118},

isbn = {979-8-3315-2627-6},

year  = {2025},

date = {2025-12-10},

urldate = {2025-12-10},

booktitle = {2025 IEEE 64th Conference on Decision and Control (CDC)},

publisher = {IEEE Control Systems Society},

address = {Rio de Janeiro, Brazil},

abstract = {This paper presents a distributed Q-learning (DQlearning) algorithm within the framework of Multi-agent Markov Decision Process characterized by heterogeneous state transition probabilities and a common reward function. By communicating during learning, each agents experiencing different state transition probabilities, do not converge to their own local optimal Q-functions. Instead, the interaction forces their learning dynamics to align, and every agent converges to the optimal Q-function corresponding to the average over these heterogeneous state transition probabilities. Further analyzing its behavior, we reformulate the update algorithm of DQlearning as a continuous time dynamics using modified ODE based stochastic approximation. Through the blended dynamics approach, the asymptotics of such dynamics is analyzed, theoretically guaranteeing the emergent behavior of DQ-learning.},

keywords = {Blended dynamics, Reinforcement learning, Stochastic Approximation},

pubstate = {published},

tppubtype = {inproceedings}

}