基于邻域模型的信号指纹特征约简
Reduction of Individual Transmitter Features Based on Neighborhood Model
为解决信号指纹的特征选择问题，基于邻域粗糙集模型生成约简规则。由粗糙集中约简和决策正域的概念出发，对由多种信号指纹组成的特征集进行约简，得到优化特征子集，并以粗糙集邻域分类器实验证明，约简后特征集维数大幅减少，且能达到不亚于甚至优于原特征集的分类效果。
o solve the problem of selecting communication transmitter identification features，we used a generation rule of rough set neighborhood relation model. Based on the concept of attribute reduction and decision positive region，a feature selection algorithm is used to reduce the data set with many kinds of communication transmitter identification features to get a reduced subset. The experiment with neighborhood classifier shows the reduced data set with much less dimensions has a similar or even better classification result with the original one.
宋家瑸、孙娜、周亚建
无线通信通信无线电设备、电信设备
粗糙集属性约简信号指纹邻域关系
rough setattribute reductionindividual communication transmitter identificationneighborhood relation
宋家瑸,孙娜,周亚建.基于邻域模型的信号指纹特征约简[EB/OL].(2010-11-09)[2025-08-16].http://www.paper.edu.cn/releasepaper/content/201011-178.点此复制
评论