A New Perspective on Algorithms for Optimizing Policies under Uncertainty

A New Perspective on Algorithms for Optimizing Policies under Uncertainty
复制标题

不确定性下优化政策算法的新视角

DOI:
--
复制
发表时间:
2000
期刊:
International Conference on Artificial Intelligence Planning Systems
影响因子:
--
通讯作者:
R. Dechter
R. Dechter
中科院分区:
--
文献类型:
--
作者:
R. Dechter

文献摘要

被引文献

相似文献

本文重新审视了一组政策的预期效用最大化的算法,也就是说,一组可能的方式来对世界的不确定状态的观察作出反应。使用桶消除框架,我们的特点,这个优化任务的复杂性,基于图形的参数,并设计一个改进的现有算法的变体。的改进,以产生一个显着的增益的复杂性时,概率子图(的影响图)是稀疏的,无论其实用程序子图引入的复杂性。
The paper takes a fresh look at algorithms for maximizing expected utility over a set of policies, that is, a set of possible ways of reacting to observations about an uncertain state of the world. Using the bucket-elimination framework, we characterize the complexity of this optimization task by graph-based parameters, and devise an improved variant of existing algorithms. The improvement is shown to yield a dramatic gain in complexity when the probabilistic subgraph (of the influence diagram) is sparse, regardless of the complexity introduced by its utility subgraph.