Applying weak equivalence of categories between partial map and pointed set against changing the condition of 2-arms bandit problem
Applying weak equivalence of categories between partial map and pointed set against changing the condition of 2-arms bandit problem
复制标题
DOI：
10.1002/cplx.20331
复制
发表时间：
2011-03
期刊：
影响因子：
--
通讯作者：
T. Niizato;Y. Gunji
中科院分区：
文献类型：
--
作者：
T. Niizato;Y. Gunji
组织学习的主要目的是使代理人在探索和利用之间取得平衡。当智能体在有限时间内确定答案时，就产生了探索与利用之间的权衡问题。这个问题从来没有出现在给定的足够的时间。当任务时间有限时，代理人必须根据自己的不完全信息来决定答案。2-在这个问题中经常考虑到武器抢劫问题。在上一篇文章中，我们提出了动态改变知识层次的弱身份模型。近年来，启发式模型的方法被吸引来解决这个问题。本文针对两臂强盗问题提出了不同的解决方法，所提出的模型能够适应环境条件的变化，为组织学习提供了另一种思路。© 2010 Wiley Periodicals，Inc. Complexity 16：10-21，2011 © 2011 Wiley Periodicals，Inc.
The main interest of the organizational learning is that the agent balances between exploring and exploiting. When the agent decides the answer in finite time, it gives rise to the problem that is trade-off between exploring and exploiting. This problem never arises in the given enough time. The agent must decide answer from his imperfect information, when the time is given finite for the task. 2-arms bandit problem is often taken into consideration in this problem. In previous article, we proposed the model for weak identity that is dynamically changing the hierarchy of his knowledge. Recently, the method of the heuristic model is attracted to approach this problem. In this article, we proposed different approaches for 2-arms bandit problem, and the model we proposed can adapt to the environment when the condition of the problem is changed and shows another approach to the organizational learning. © 2010 Wiley Periodicals, Inc. Complexity 16: 10-21, 2011 © 2011 Wiley Periodicals, Inc.