首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于沃尔什变换的连续型数据离散化方法及应用
引用本文:王晓芳,王泽华,黄龙,陈莉.基于沃尔什变换的连续型数据离散化方法及应用[J].新疆职业大学学报,2005,13(1):57-59,42.
作者姓名:王晓芳  王泽华  黄龙  陈莉
作者单位:1. 西南石油学院,四川,成都,610059;石油大学,北京,102249
2. 新疆石油学院,新疆,乌鲁木齐,830000
3. 石油大学,北京,102249
基金项目:中国石油天然气集团公司中青年科技创新基金 (编号 :CX2 0 0 0 - 14 )
摘    要:在智能信息处理中,大多要求整型或字符型的离散化数据,而实际数据往往是连续型的。本文在沃尔什变换的基础上,提出一种连续数据离散化方法。该方法不需要给定划分维数,连续数据经过沃尔什低通滤波后被划分成离散的区间,再用整数和符号标示。通过实际数据测试,该方法取得了较好的效果。

关 键 词:数据挖掘  离散化  沃尔什变换  连续型数据  人工智能
文章编号:1009-9549(2005)01-0057-03

A Method of Continuous Data Discretization Based on Walsh Transformation and It's Application
Wang Xiaofang,Wang Zehua,Huang long,Chen li.A Method of Continuous Data Discretization Based on Walsh Transformation and It''''s Application[J].Journal of Xinjiang Vocational University,2005,13(1):57-59,42.
Authors:Wang Xiaofang  Wang Zehua  Huang long  Chen li
Institution:Wang Xiaofang 1,3,Wang Zehua 2,Huang long 3,Chen li 3
Abstract:In intelligent information system, integer data or character data are often required, but real data is usually continuous. To resolve this problem a method is presented to discretize data, which is based on Walsh transformation.The dimensions are not needed in this method. Continuous data are divided into discrete ranges by means of low-pass-filter and the data are labeled with integer and symbol data. Through practical testing with real data, the algorithm is very efficient.
Keywords:Intelligent System  Discretization  Walsh transformation
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号