权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Optimal Allocation Problem of Social Capital through a Recursive Learning System and a Multistage Decision Making

通过递归学习系统和多级决策的社会资本优化配置问题

基本信息

批准号：
07680468
负责人：
NAKAI Toru
金额：
$ 1.28万
依托单位：
KYUSHU UNIVERSITY
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
1995
资助国家：
日本
起止时间：
1995 至 1996
项目状态：
已结题

项目摘要

In the first year 1995 of our project, we focused our activity to the study to analyze allocation problems of the social capital as a sequential decision problem. We also considered processes to induce a recursive method as a learning procedure. Furthermore, in the context of the sequential allocation problem, we made a research from the point of the dynamic programming. The results were presented in several workshops, and the analytical comments were summarized in some papers.As a result, it was found that the theory of the dynamical system is effective to analyze these problems having complex features and factors such as an allocation problem of the social capital. Moreover, we have a perspective that the optimization methods for a partially observable decision process are useful to consider a role of a learning procedure which is tightly combined with recursive processes for sequential allocation problems.In the second year 1996, we made our project more fruitfully by collecting an amount of documents applied to our studies. We also studied about some fields related to our project, and obtained analytical and technical results. In terms of the theoretical contribution of out project, we published several works in 1995 and 1996 in the fields of the basic theory of the dynamic programming, the inventory theory, the optimal policy for stochastic systems, applications of time series, multimedia environment and the networking system. Furthermore, in relation to the necessity to analyze a learning procedure, we published some papers on the problem for stochastic order relations for a recursive system.

在1995年项目的第一年，我们将研究的重点放在了将社会资本的分配问题作为一个序列决策问题进行分析的研究上。我们还考虑了诱导递归方法的过程作为学习过程。此外，针对序贯分配问题，从动态规划的角度对其进行了研究。结果表明，动力系统理论能够有效地分析这些具有复杂特征和因素的问题，如社会资本的配置问题。此外，我们的观点是，部分可观测决策过程的优化方法有助于考虑与递归过程紧密结合的学习过程对序列分配问题的作用。在1996年的第二年，我们通过收集大量应用于我们的研究的文献，使我们的项目更有成效。我们还对与本项目相关的一些领域进行了研究，并取得了分析和技术成果。根据OUT项目的理论贡献，我们于1995年和1996年在动态规划的基础理论、库存理论、随机系统的最优策略、时间序列的应用、多媒体环境和网络系统等领域发表了几项工作。此外，关于分析学习过程的必要性，我们发表了一些关于递归系统的随机序关系问题的论文。