摘要
用模糊似然函数计算类内及类间距离，得到任意特征子集的模糊特征选择系数，用于特征子集的选择，从而得出最能区分和表征模式类的特征子集．举例说明了该方法的具体用法，表明具有好的实用性．
The fuzzy feature selection index are defined using the distances of intraset and interset of pattern classes,which are calculated by fuzzy likelihood function. The algorithm can be used to compare the importance of any subset of features,so that the most important feature subset can be gotten. An example indicates how to use them in concrete situations,which shows that these methods are applicable.
出处
《控制理论与应用》
EI
CAS
CSCD
北大核心
1998年第4期646-648,共3页
Control Theory & Applications
基金
国家自然科学基金!69334012
关键词
模糊特征选择
模糊似然函数
模式识别
fuzzy feature selection
fuzzy likelihood function
feature selection index
