|
12月24日芬兰VTT技术研究中心Xianfu Chen博士学术报告预告
作者:控制学科
发布日期:2018-12-24
浏览次数:
时间:12月24日下午13:30-14:30 报告摘要:Markov decision process (MDP) is a promising mathematical framework to model autonomous behaviours in an uncertain networking environment. In this talk, we will first introduce basics of a MDP. Without a priori knowledge of the environmental dynamics statistics, reinforcement learning algorithms make sense in such scenarios and are easy for implementation. Through several use case studies from our most recent efforts, we will see how reinforcement learning algorithms achieve the desired objectives by exploiting limited information of feedbacks received from the environment. |
