Risk-Averse Learning by Temporal Difference Methods

Risk-Averse Learning by Temporal Difference Methods
复制标题

通过时间差异法进行风险规避学习

DOI:
--
复制
发表时间:
2020
影响因子:
6
通讯作者:
Ümit Emre Köse
Ümit Emre Köse
中科院分区:
计算机科学3区
文献类型:
--
作者:
Ümit Emre Köse

文献摘要

被引文献

相似文献

我们考虑强化学习,其性能由动态风险度量评估。构造了一个投影风险规避动态规划方程,并研究了它的性质。然后,我们提出了风险厌恶的时间差异的方法,我们证明了他们的收敛概率为1。我们还进行了一个复杂的运输问题的实证研究。
We consider reinforcement learning with performance evaluated by a dynamic risk measure. We construct a projected risk-averse dynamic programming equation and study its properties. Then we propose risk-averse counterparts of the methods of temporal differences and we prove their convergence with probability one. We also perform an empirical study on a complex transportation problem.