决策信息系统分解约简研究论文.docVIP

摘 要 粗糙集是一种处理不精确、不一致和不完备信息的数学工具，但在实际应用中，数据规模庞大、结构复杂，限制了其应用，在数据分析之前对决策信息系统进行分解是一种有效的数据转换方法。本文主要研究决策信息系统分解约简方法。首先介绍粗糙集理论的基础知识，包括等价关系、信息系统和决策系统、上下近似、约简与核等概念。然后详细论述比较了两种决策信息系统分解方法，即基于粗糙集属性度量的决策信息系统分解方法和属性层次分解约简方法，通过分析，说明了两种方法具有信息量无损，时间复杂度低，实用性强，动态特性好等优点，文中对两种方法均给出了实例说明。根据基于粗糙集属性度量的分解方法，设计并实现了一个分解约简实验系统，详细论述了系统的目标及功能，主要模块的算法思想，介绍系统的实现过程，包括开发平台、编程工具、数据源、数据库等的选取。通过对实验数据的测试验证了该分解约简方法的可行性。最后，对全文进行了概括性总结，并提出了有待进一步研究和完善的问题。 关键词 粗糙集，决策信息系统，分解，约简 ABSTRACT Rough set theory is a new mathematical tool for processing imprecise, inconsistent and incomplete information, but the traditional rough set method becomes inefficient while dealing with large data sets. Decomposition of the large decision information system is an effective approach to solve the problem. The thesis focuses on decomposition and reduction approaches of decision information system. Firstly, the basic knowledge of rough set theory is introduced in detail, including equivalent relations, information and decision system, upper/lower approximation, core, reduction, etc. Secondly, two different decomposition and reduction approaches of decision information system are analyzed and compared in the thesis. The introductory examples to the approaches are given. It is proved that the two approaches will not cause the loss of information, while computing complexity decreases at the same time. Subsequently, according to the decomposition approach based on rough set attribute measurement, an experimented system is designed and implemented. The aim of the system is presented and the main functional modules as well as algorithms are described in detail. The testing result on the experimented data set shows the applicability of the approach to practice. Finally, a recapitulative conclusion is given, and the problems for further research are pointed out. KEY WORDS rough set, decision information system, decomposition, reduction 目 录 摘 要 I ABSTRACT II 第一章 绪 论 1 1.1 研究背景 1 1.2 国内外研究现状 2 1.2.1 粗糙集理论研究现状 2 1.2.2 决策信息系统分解约简研究现状 3 1.3 论文