arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1904.09024cs.LGcs.AIstat.ML
收藏

什么时候有预测知识?

When is a Prediction Knowledge?

University of Alberta(阿尔伯塔大学)

Alex Kearney, Patrick M. Pilarski

摘要:

在强化学习中,越来越多的研究旨在通过对感觉、行为和时间的预测来表达智能体对世界的所有知识。这项工作不仅可以被看作是一个架构建议的集合,也可以被看作是强化学习中机器知识理论的开端。最近的工作扩展了可以使用预测表达的内容,并开发了使用预测来为各种合成和现实问题的决策提供信息的应用程序。虽然很有希望,但我们在这里认为,在强化学习中,预测作为知识的概念还不成熟:一些工作明确地将预测称为知识,将预测视为知识的要求还有待深入探讨。这种对知识的必要和充分条件的规范是重要的;即使关于知识性质的主张隐含在技术提案中,这种主张的基本假设也会对我们设计的系统产生影响。这些后果体现在我们选择构建预测知识架构的方式以及我们如何评估它们。在本文中,我们采取的第一步,形式化的预测知识,通过讨论的关系,预测知识的学习方法,现有的知识理论在认识论。具体而言,我们探讨广义价值函数和认识论的理由和真理的概念之间的关系。

英文摘要:

Within Reinforcement Learning, there is a growing collection of research which aims to express all of an agent's knowledge of the world through predictions about sensation, behaviour, and time. This work can be seen not only as a collection of architectural proposals, but also as the beginnings of a theory of machine knowledge in reinforcement learning. Recent work has expanded what can be expressed using predictions, and developed applications which use predictions to inform decision-making on a variety of synthetic and real-world problems. While promising, we here suggest that the notion of predictions as knowledge in reinforcement learning is as yet underdeveloped: some work explicitly refers to predictions as knowledge, what the requirements are for considering a prediction to be knowledge have yet to be well explored. This specification of the necessary and sufficient conditions of knowledge is important; even if claims about the nature of knowledge are left implicit in technical proposals, the underlying assumptions of such claims have consequences for the systems we design. These consequences manifest in both the way we choose to structure predictive knowledge architectures, and how we evaluate them. In this paper, we take a first step to formalizing predictive knowledge by discussing the relationship of predictive knowledge learning methods to existing theories of knowledge in epistemology. Specifically, we explore the relationships between Generalized Value Functions and epistemic notions of Justification and Truth.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑