• 主题
高级搜索  >
历史搜索 清空历史
首页> 中文期刊> 《小型微型计算机系统》 >一种基于互信息的实时特征提取算法

一种基于互信息的实时特征提取算法

         
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

在工业生产过程中由于数据采集设备多样,采样频率高,各变量之间相互耦合,导致工业大数据维度高、数据间呈现出非线性关系、处理实时性要求高.传统的特征提取方法会降低非线性数据在后续数据挖掘中的精度,同时其数据处理速度无法满足工业大数据的实时性要求.针对该问题,提出一种基于互信息的实时特征提取算法(MIFE);采用改进的互信息作为提取特征的评价准则,以提高非线性数据在后续挖掘中的精度;同时采用自适应的滑动窗口技术以较小代价实现对增量数据的实时处理,将历史数据和动态新增数据结合,实现了整个数据集的快速降维.实验结果表明MIFE算法可以快速地对数据进行特征提取,并在多种分类器上验证了算法的准确率.%Due to the diversity of data acquisition equipment and the high sampling frequency in the industrial production process,each variable is coupled to each other,which results in high dimensions of industrial big data,strong internal correlation of data,and high re-quirements for real-time processing. Traditional feature extraction methods reduce the accuracy of nonlinear data in subsequent data mining. And its data processing speed can not meet the real-time requirements of industrial big data. Aiming at this problem,a real-time feature extraction algorithm based on mutual information(MIFE)is proposed. Mutual information is used as an evaluation criteri-on for extracting features to improve the accuracy of subsequent data mining in nonlinear data. At the same time,the sliding window technology is used to achieve a lower cost in real-time processing of incremental data,historical data and new data are combined to quickly achieve dimension reduction of the entire sample data set. Experimental results show that the MIFE algorithm can quickly ex-tract features from the data and verify the accuracy of the algorithm with multiple classifiers.

著录项

相似文献

  • 中文文献
  • 外文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号