%0 Journal Article
%T A Method of Feature Selection and Classification Based on Divergence<br>基于分离度的图象特征提取与识别方法
%A Lian ShiZhu
%A <br>连石柱
%J 中国图象图形学报
%D 1996
%I 
%X The separability of two pattern classes can be measured by the divergence for two Gaussian distributions. Because the divergence is invariant under the linear transformation we can extract "good" features for separating two patterns via Karhunen-Loeve transformation. It is shown that the divergence is only dependent on two of n eigenvalues. One property of a "good" dichotomy is that each feature should be effective for classification. Thus, a criterion function is proposed. This algorithm, which is called as sample exchange algorithm, is convergent and it is a reasonable unsupervised clustering method for classification.
%K Feature selection
%K Mixture normalization
%K Sample autocorrelation matrix
%K K-L expansion
%K Eigenvalue
%K Eigenvector
%K Divergence
%K Iteration
%K Clustering<br>特征提取
%K 混合归一化
%K 样本自相关阵
%K 图象识别
%U http://www.alljournals.cn/get_abstract_url.aspx?pcid=5B3AB970F71A803DEACDC0559115BFCF0A068CD97DD29835&cid=8240383F08CE46C8B05036380D75B607&jid=D06194629680C940ACE75262F54B9D85&aid=A514751AF0BC9534A73B9118F8AC9201&yid=8A15F8B0AA0E5323&vid=CA4FD0336C81A37A&iid=38B194292C032A66&sid=E0F6F365E4766526&eid=1E41DF9426604740&journal_id=1006-8961&journal_name=中国图象图形学报&referenced_num=0&reference_num=0