BriefGPT.xyz
Dec, 2017
通过非平衡热力学对图上最优分层策略推断进行表征
Characterizing optimal hierarchical policy inference on graphs via non-equilibrium thermodynamics
HTML
PDF
Daniel McNamee
TL;DR
该论文介绍了一种新的推断方法来构建状态空间层次结构,从而得到一种层次化的策略推断算法,用以逼近先前和最优策略之间在状态空间轨迹密度上的离散梯度流。
Abstract
hierarchies
are of fundamental interest in both
stochastic optimal control
and
biological control
due to their facilitation of a range of
→