| 基于模糊分裂的概念自适应增量决策树分类算法 |
| |
| 引用本文： | 朱参世,张超,李响.基于模糊分裂的概念自适应增量决策树分类算法[J].计算机工程与设计,2011,32(4):1414-1418. |
| |
| 作者姓名： | 朱参世 张超 李响 |
| |
| 作者单位： | 1. 空军工程大学,工程学院,陕西,西安,710038
2. 空军工程大学,导弹学院,陕西,西安,713800 |
| |
| 摘 要： | 针对数据流环境下传统分类挖掘算法的不足,引入了改进的滑动窗口技术和模糊技术,通过在滑动窗口中设置分类效用因子的方法提高了窗口的利用率,有效改善了由于概念漂移所带来的分类器过时的问题,在连续属性分裂过程中加入了模糊技术,解决了连续属性字段平滑离散化的问题。理论分析和实例表明了改进后的算法具有较低的运行环境要求和较高的分类准确率。
|
| 关 键 词： | 数据流 分类 决策树 滑动窗口 模糊离散化 |
Concept adaptive incremental decision tree assorted arithmetic based on fuzzy division |
| |
| ZHU Can-shi,ZHANG Chao,LI Xiang.Concept adaptive incremental decision tree assorted arithmetic based on fuzzy division[J].Computer Engineering and Design,2011,32(4):1414-1418. |
| |
| Authors: | ZHU Can-shi ZHANG Chao LI Xiang |
| |
| Affiliation: | 1.College of Engineering,Air Force Engineering University,Xia’n 710038,China;2.College of Missile,Air Force Engineering University,Xia’n 713800,China) |
| |
| Abstract: | In accordance with some defects of custom classification mining arithmetic in data stream environment,modified sliding window technique and fuzzy theory is introduced.By using classify factor in the window,the utilization ratio of it had been raised which makes the outdate problem of the classification improved.By adding fuzzy theory into the continuous attributes split procedure,we suc-cessfully solved a problem that changes continuous attributes into discrete ones smoothly.The approach is demonstrated to ... |
| |
| Keywords: | data stream classification decision tree sliding window fuzzy discrete |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
