基于加权Bayes分类算法的不完备信息系统数据挖掘研究  

Reseach for Data Mining of Incomplete Information Systems Based on Weighted Naive Classification Algorithm

作　　者：李莉[1] 赵晋强[2] LI Li, ZHAO Jin-qiang2（1 .China Institute of Defence Science and Technology , College of Arts ＆ Science,Beijing 101601,China; 2. PLA Headquart of the General Staff Army Aviation Research Institute, Beijing 101114,China）

机构地区：[1]中国防卫科技学院文理学院信息工程系,北京101601 [2]总参谋部陆航研究所,北京101114

出　　处：《电脑知识与技术》2007年第9期1408-1409,1480,共3页Computer Knowledge and Technology

基　　金：教育部科学技术研究项目（02036）

摘　　要：基于相似粗集理论模型,对加权朴素Bayes算法进行了扩展,同时改进了传统不完备信息系统中缺失信息的弥补方法,并由此提出了基于不完备信息系统的加权Bayes分类算法,阐述了其对于不完备系统数据挖掘的重要意义,通过计算机仿真实验验证了该方法的有效性。Weighted Na？ve Classification Algorithm is extended based on Comparability Rough sets theory. The original recuperation method of lost information in Incomplete Information Systems is improved too. Weighted Na？ve Classification Algorithm based on Incomplete Information Systems is developed, and its slgmficance for data mining is also set forth. Simulation results on a variety of data set illustrate the efficiency of this new algorithm.

关 键 词：粗集理论 加权朴素Bayes 不完备信息系统 数据挖掘 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论;自动化与计算机技术—计算机科学与技术]