Markov decision processes under ambiguity
Markov decision processes under ambiguity
复制标题
模糊条件下的马尔可夫决策过程
DOI：
10.4064/bc122-2
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
U. Rieder
中科院分区：
文献类型：
--
作者：
Nicole Bauerle;U. Rieder
我们考虑统计马尔可夫决策过程，其中决策者厌恶模型模糊性的风险。后者由影响转移定律和成本函数的未知参数给出。风险厌恶程度可以通过熵风险度量或平均风险价值来衡量。我们展示了如何使用一般的极小极大定理来解决此类问题。在一些连续性和紧凑性假设下，我们证明最优（确定性）策略的存在并讨论其计算。我们使用统计决策理论的例子来说明我们的结果。
We consider statistical Markov Decision Processes where the decision maker is risk averse against model ambiguity. The latter is given by an unknown parameter which influences the transition law and the cost functions. Risk aversion is either measured by the entropic risk measure or by the Average Value at Risk. We show how to solve these kind of problems using a general minimax theorem. Under some continuity and compactness assumptions we prove the existence of an optimal (deterministic) policy and discuss its computation. We illustrate our results using an example from statistical decision theory.