【作者】 罗艺；

【导师】 陈炼；

【作者基本信息】 南昌大学 ， 计算机软件与理论， 2009， 硕士

【摘要】 在信息时代,存在着大量的信息集合,人们可以从这些信息中挖掘有用的信息,但是实际中大多数这样的信息系统的数据具有不确定性。随着在实际问题中处理不确定性集合方法的不断发展,粒计算理论成为世界范围内一种常用的研究方法和应用广泛的处理不确定性的方法。粒计算现已成为人工智能领域研究的热点之一。本文是基于粒计算理论的信息熵等相关知识对非完备决策表进行约简的研究,并对比分析了该非完备决策表信息更新前后信息熵及约简的情况。本文主要工作如下:(1)概述了粒计算基本理论、研究趋势和发展现状;(2)介绍了非完备决策表、粒计算理论中的信息熵、信息粒度等定义;(3)在完备信息系统基于信息熵的约简算法基础上,给出了非完备决策表基于信息熵的约简算法;(4)对比分析了非完备决策表数据更新前后信息熵、约简过程和约简结果的变化。得出信息熵值会发生变化,信息熵值的降序排列也会发生变化,并且对于非海量信息的信息系统动态数据更新是必要的。更多还原

【Abstract】 In the information age, there are a lot of assemble of information in the world. People can dig out a lot of useful information from these assembles, but in the real some data from information systems is uncertainty. Develping of the methodologies for dealing with uncertainty in real-world problems, granular computing is becoming a widespread and useful method in the worldwide research community and exhibiting a broad range of applications. Granular computing now has become one of hot spots in artificial intelligence.In this article, the research of reduction in the incomplete decision table base on the granular computing theory. It also analyse the data of information entropy and the result of the reduction. The main works are as follows.(1) The article gives a brief introduction about granular computing and its development, current research;(2) It introduces the definition about the incomplete decision table and information entropy, information granularity in granular computing theory and so on;(3) Basing on the reduction algorithm using information entropy theory in the complete information systems, the article gives the other reduction algorithm in the incomplete decision table;(4) Contrast the date of information entropy, the procedure for reduction and the result of reduction after data update. We can know that the information entropy and the sequence of the decreasing order about the data of information entropy would change. We also can know that when the data set of the information systems is not very big, it is necessary to update the information in time.更多还原