基于S粗集的系统规律挖掘与非线性系统输出反馈-概率论与数理统计专业论文.docxVIP

11 1111 IIl 1 76 学习、数据挖掘等领域得到广泛的应用．应用粗集理论解决上述问题，最成熟的 方法实质都可归结为：规则提取娜3：Z一粗糙集是具有静态特性的集合x的粗集， 因此其应用范围是十分有限的，这使得粗集理论在应用上陷于瓶颈，迫切需要新 理论、新方法的出现．特别是对于动荡的金融系统、投资系统、智能系统、目标 多变的识别一分析系统、动态数据挖掘、动态知识发现的研究遇到困难．2002年， 山东大学的史开泉教授提出了S一粗集理论，他将粗集理论从静态推广到动态， 提出了规律挖掘的思想，S一粗集是以Z-Pawlak粗集理论为基础，对Z-Pawlak粗 集理论拓广和延伸． 本文通过对S一粗集、函数S一粗集的讨论，为变化多端的金融系统、风险 投资系统、系统状态识别、故障诊断系统统等诸多系统中的规律识别扭一棚1，为 系统状态规律生成及规律挖掘、规律发现的研究提供新的理论支持和新的研究思 想但】．本文对S一粗集、函数S一粗集结构及应用作进一步的的探讨，对动态的系 统规律给出了挖掘方法，用样条插值函数来发现动态的规律：给出了S一粗集在 网络数据中的分层挖掘，医疗诊断系统、地震预测系统中的应用；另外本文还对 非线性系统的输出反馈做了研究，给出了一类非线性系统有限时间稳定和稳定输 出跟踪． 关键词：S一粗集；规律生成；非线性系统；输出反馈 Abstract In 1 982，Rough set theory proposed by Z．Pawlak is a tool which can deal with integrity and uncertainty mathematical．It can analysis and deal with Various kinds of implicit information which is inaccurate， consistent and incomplete and etc and reveal the potential law．The main idea is finding the problems decision or classification rules by knowledge reduction in the condition of keeping the classification ability stable and getting extremely extensive application in pattern recognition，tmehine learning and data mining and ete．The best way to solve the above problems by Z-rough set is rule extract．Z·-rough set is a set of static chamcteristics．Therefore its application is very limited and a new method is dermnded urgently，especially for volatile financial system，investment system, intelligent system，target changeful recognition一锄lysis system,dynamic data mining，dynamic knowledge discovery．In 2002，Shi Kaiquan，the professor of Shandong university put forward S—rough set rough theory．He developed rough set theory from the static promotion to dynamic and put forward the thoughts of rule mining．S-rough set is the extension ofZ—rough which is a foundation of S—rough set． This article support a new theory and thought for the hw recognition of the protean financial system，risk investment system，state identification system,fault diagnosis system and SO on and law generation and law minin