Abstract
We propose a multi-agent reinforcement learning model, Graphon Game with Re-Sampling (GGR-S), for large but finite population dynamics. We introduce a re-sampling scheme where agents re-sample their connections at each time step from a piece-wise constant graphon. We analyze the GGR-S dynamics and establish its convergence properties. Moreover, we propose efficient locally centralized and decentralized sample-based N -agent Reinforcement Learning algorithms for GGR-S and provide rigorous convergence analyses with finite sample guarantee. (joint with Peihan Huo, Oscar Peralta, Junyu Guo & Qiaomin Xie)
Bio:
Andreea Minca is a Professor in the School of Operations Research and Information Engineering at Cornell University. She holds degrees from Sorbonne University (PhD in Applied Mathematics) and Ecole Polytechnique (Diplome de l'Ecole Polytechnique).
In recognition of "her fundamental research contributions to the understanding of financial instability, quantifying and managing systemic risk, and the control of interbank contagion", Andreea received the 2016 SIAM Activity Group on Financial Mathematics and Engineering Early Career Prize. Andreea is also a recipient of the NSF CAREER Award (2017), a Research Fellow of the Global Association of Risk Professionals (GARP) (2014), and an AXA Research Fund Awardee (2020).
Join Zoom Meeting
https://us06web.zoom.us/j/84940856266?pwd=KnBqwY5RuDTrwbWPLumDNtPERgU9zC.1
Meeting ID: 849 4085 6266
Passcode: 737702
