报告题目：Kernel-based Decentralized Policy Evaluation for Reinforcement Learning

报告人：练恒教授 香港城市大学

报告时间：2024年5月14日（星期二），上午10:00

报告地点：兴庆校区数学楼2-1会议室

报告摘要：We investigate the decentralized nonparametric policy evaluation problem within reinforcement learning, focusing on scenarios where multiple agents collaborate to learn the state-value function using sampled state transitions and privately observed rewards. Our approach centers on a regression-based multi-stage iteration technique employing infinite-dimensional gradient descent within a reproducing kernel Hilbert space (RKHS). To make computation and communication more feasible, we employ Nystrom approximation to project this space into a finite-dimensional one. We establish statistical error bounds to describe the convergence of value function estimation, marking the first instance of such analysis within a fully decentralized nonparametric framework. We compare the regression-based method to the kernel temporal difference (TD) method in some numerical studies.

个人简介：练恒，现任香港城市大学数学系教授，于2000年在中国科学技术大学获得数学和计算机学士学位，2007年在美国布朗大学获得计算机硕士，经济学硕士和应用数学博士学位。先后在新加坡南洋理工大学，澳大利亚新南威尔士大学，和香港城市大学工作。在高水平国际期刊上发表学术论文30多篇，包括《Annals of Statistics》《Journal of the Royal Statistical Society，Series B》、《Journal of the American Statistical Association》《Journal of Machine Learning Research》《IEEE Transactions on Pattern Analysis and Machine Intelligence》. 研究方向包括高维数据分析，函数数据分析，机器学习等。