一种改进的最佳鉴别平面  

An Improved Optimal Discriminant Plane

作　　者：赵海涛[1] 金忠[1] 

机构地区：[1]南京理工大学理学院,南京210094

出　　处：《南京理工大学学报》2000年第1期88-92,共5页Journal of Nanjing University of Science and Technology

摘　　要：在模式识别领域 ,基于Fisher鉴别准则函数的Sammon最佳鉴别平面技术有着重大的影响。特征抽取的一般原则是最好抽取模式的不相关的特征 ,而Sam mon最佳鉴别平面的 2个鉴别投影方向虽然是正交的 ,但一般也是统计相关的。金忠等[1] 提出了一种统计不相关的最佳鉴别平面 ,但其 2个鉴别投影方向却一般不是正交的。文中从Fisher鉴别准则函数出发 ,提出了一种改进的最佳鉴别平面 ,并在最后给出了改进的最佳鉴别矢量集的求解方法。实验结果表明 。Based on Fisher's discriminant function,Sammon's optimal discriminant plane has great influence in the area of pattern recognition.A general rule for feature extraction is to extract features as uncorrelatedly as possible,but the two features in Sammon's optimal discriminant plane are correlated.A different optimal discriminant plane was presented before.Although the two features in this plane are uncorrelated,they are not orthogonal.Based on the improvement of Fisher's discriminant function, an optimal discriminant plane with uncorrelated and orthogonal features is presented in this paper.Experiments on a three class problem are performed.Experimental results have shown that the new optimal discriminant plane is very effective.

关 键 词：模式识别 特征抽取 最佳鉴别平面 特征选择 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] O235[自动化与计算机技术—计算机科学与技术]