CASIA OpenIR  > 毕业生  > 硕士学位论文
智能化经济决策支持系统研究与神经元网络在其中的应用
李京山
Subtype工学硕士
Thesis Advisor郑应平
1992
Degree Grantor中国科学院自动化研究所
Place of Conferral中国科学院自动化研究所
Degree Discipline控制理论与控制工程
Abstract决策支持系统是计算机信息系统发展的新阶段,是计算机科学,管理 决策科学,系统控制理论与行为科学等相结合的产物,是计算机信息系统 的发展方向。其中,智能化决策支持系统代表着DSS研究与应用的主流, 具有广阔的发展前景。开发宏观经济智能决策支持系统不论是对于决策支 持系统本身的研究,还是对于科学的决策和宏观经济理论的研究,都有着 十分重要的意义。本文重点研究了智能经济决策支持系统的设计思想和人 工神经元网络在其中的应用。 在经济过程中应用智能决策支持系统将是对更科学有效的经济决策的 一种挑战和帮助。计算机信息系统的发展经历了电子数据处理,管理信息 系统,专家系统等阶段,在七十年代发展出决策支持系统,并在世界范围 引起了一股研究直用热潮。计算机科学的发展是决策支持系统发展的动力 和基础,智能化是决策支持系统发展的方向。作者认为:决策支持系统是 支持决策者决策过程的,主要用来解决半结构化和非结构化问题的智能化 ,综合性计算机信息系统。经济决策支持系统研究是我国经济改革取发展 的需要。应用综合集成方法的智能经济决策支持系统将有助于为经济决镱 者提供政策咨询,并有助于将DS$研究提高到新的高度。DSS的发展造切要 求DSS的智能化,同时要求综合性的DSS。在研究过程中,经过不断总结与 讨论,报据宏观经济决策的特点,作者提出了智能化,综合性决策支持系 统的构想,指出它将是具有较高智能和较强综合能力的,以知识库系统为 核心与基础的新一代DSS。同时,对ICDSS还提出了基于知识的结构框架, 它以知识库为基础,以知识管理系统统一协调管理整个系统。分析了其系 统结构,并重点讨论了其知识库系统的内容和结构。决策支持系统要面向 决策过程,通过决策者与计算机在决策过程各阶段的交互来支持决策。通 过对宏观经济决策过程的分析,对于实现面向决策过程的DSS设计与开发 有着重要的意义。八十年代中期开始,国际上掀起了一股人工神经元网络 研究的热潮,亲自不同领域的众多科学工作者参加了这一竞赛。神经网络 系统是大量神经元广泛相互连接而成的复杂网络系统。BP网络是神经元网 络研究中的重要模型之一。它通过将误差反向传播来调整权重,最终实现 网络学习。随着人工神经元网络研究热潮的推动,在经济决策过程中应用 神经元网络是一种新的尝试。通过学习使网络对一类特定输入输出模式产 生敏感性,来实现对类似决策过程的支持。有益于决策支持的智能化和定
Other AbstractDecision Support System (DSS) is the new phase of Computer-Base information System (CBIS) development. It is the combination of Computer Science, Management & Decision Science, System Control Theory and Behavior Science, etc. Where Intelligent Decision Support System (IDSS) is the main direction of DSS research and application. The development of macro-economy IDSS has important meaning to the DSS research, scientific decision-making and research of macro-economy theory. This paper mainly discusses the development of Intelligent Economic Decision Support System (IEDSS) and the application of Neural Network (NN) in IEDSS. The application of IDSS in economic process will be a challenge and aid to more scientific and efficent economic decision-making. From 1970's, there has been a worldwide mass fervor of DSS research and application. The development of computer science is the base and motivation of DSS development, and intelligence is the direction of it. I think that DSS should be intelligent comprehensive CBIS, which should support decision process of decision-maker, and mainly be used to solve semi- and un-structured problems. Economic DSS research is the need of economic reform and development of China, and IEDSS will be useful to provide policy advice for economic decision-makers. The development of DSS is necessary to explore intelligent DSS and comprehensive DSS. Therefore I present a new idea of Intelligent Comprehensive Decision Support System (ICDSS). It has higher intelligence and stronger comprehensive ability. Knowledge Base System is the kernel and base of TCDSS, and the whole system is coordinated and managed by Knowledge Management System. Also I present a knur l edge-based framework of ICDSS.DSS should be decision process oriented,and be used to support decision-making through the interaction of decision-maker and computer. From mid。1980's,neural network research has been paid closed attention by worldwide scientists.Neural Network system is a complex network system with a large number neurons distributedly connected each other.Back Propagation(BP)net is one of important model in NN research.The application of NN in economic decision process is a new effort.It implements support to simi-liar decision process by making the net sensitive to a kind of input-output pattern through learning.This paper improves the learning alg07ithm of BP net to make it has better convergence rate,and discusses the parameter selection, sample effectiveness, hidden unit selection,etc.in BP net.Also it intends to apply Newton's method in NN and introduces the training algorithms of Newton method.The applications of NN in economic decision process, such as macro-economy evaluation。 business cycle analysis and prediction.to support decision-making。have got better results. The research and application of IEDSS will promote the development of DSS
shelfnumXWLW241
Other Identifier241
Language中文
Document Type学位论文
Identifierhttp://ir.ia.ac.cn/handle/173211/7032
Collection毕业生_硕士学位论文
Recommended Citation
GB/T 7714
李京山. 智能化经济决策支持系统研究与神经元网络在其中的应用[D]. 中国科学院自动化研究所. 中国科学院自动化研究所,1992.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[李京山]'s Articles
Baidu academic
Similar articles in Baidu academic
[李京山]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[李京山]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.