（毕业设计论文）《异常数据挖掘研究》.doc

PAGE 毕 业 设 计（论文） 中文题目 异常数据挖掘研究 英文题目 Research on outlier data mining 系 别： 计算机科学与技术系 专业年级： 计算机科学与技术09级 姓 名： 学 号： 0907045143 指导教师： 职 称： 讲师 2011 年 5 月 20 日 摘 要 粗糙集理论，它是一种分析处理数据的理论，在20世纪80年代由波兰科学家Pawlak建立。一开始由于语言交流上的问题，建立该理论的时候只有一些东欧学者会研究和应用它，后来随着该理论的发展才慢慢受到全球上知名数学学者和计算机学者的重视。 知识粒度的基本思想在许多领域都有体现，如粗糙集、数据库、聚类分析、模糊集、证据理论、数据挖掘和机器学习等。自从1979年L.A.Zadeh在世界上首次提出并讨论了知识粒度问题之后，知识粒度获得了人们越来越多的关注。经过多年的发展，知识粒度已在知识发现、数据挖掘、软计算中扮演越来越重要的角色。 在本论文中，本人引入知识粒度这一个方法作为一个统一的框架去理解和实施异常点挖掘。此外，还给出了基于知识粒度的异常点挖掘算法。该算法结合粗糙集与数据挖掘技术研究异常数据。理论研究和实验结果表明，知识粒度方法对评定异常是有效且适用的。 关键词：粗糙集；异常检测；数据挖掘；知识粒度 ABSTRACT Rough set theory, it is a theory of data analysis and processing, 80 in the 20th century by Polish scientists Pawlak established. The beginning of the problems due to language exchanges, the establishment of the theory, when only a few Eastern European scholars would study and apply it, and later with the development of the theory slowly by the global computer on the well-known scholars and scholars in mathematics seriously. The basic idea of ??knowledge granulation in many areas, such as rough sets, database, clustering analysis, fuzzy sets, evidence theory, data mining and machine learning.1979 L.A.Zadeh the first time in the world and discussed after the knowledge granulation of fuzzy, the knowledge granulation obtained people more and more attention.After years of development, knowledge granularity in knowledge discovery, data mining, soft computing plays an increasingly important role. In this paper, we introduce this method of knowledge granularity as a unified framework to understand and implement outlier mining. In addition, we also give the size of knowledge-based algorithm for mining outliers. The algorithm combines rough sets and data mining of abnormal data. Theoretical and experimental results show that the method of assessment of knowledge granularity is an effective and appropriate exception. Key Words: rough sets; outlier detection; data mining; knowledge granulation 目