本文提出一种利用平行坐标图的多元信息表示对主成分分析特征提取方法进行优化的分类技术。
A novel method for optimizing the principle component analysis in feature extraction is proposed, which making use of parallel coordinate plot for graphical presentation of multivariate information.
在此基础上，采用主成分分析法对“五因素”进行特征提取，降低BP网络的输入维度。
On the above basis, we used principal component analysis of the "five factors" for feature extraction and reduced the input dimension of BP network importation.
在图像特征提取上改进并提出了三种特征的提取：纹理特征，灰度直方图均值化特征，图像的主成分特征。
The features concerned are such as texture feature, gray histogram feature and features derive form the principle component analysis.
通过实验比对可知该算法效果在特征提取和分类方面均优于传统核主成分分析法以及最近邻分类器。
The experimental comparisons show that this algorithm outperforms traditional KPCA and K-Nearest Neighbor classifier on both feature extraction and classification.
通过实验比对可知该算法效果在特征提取和分类方面均优于传统核主成分分析法以及最近邻分类器。
The experimental comparisons show that this algorithm outperforms traditional KPCA and K-Nearest Neighbor classifier on both feature extraction and classification.
应用推荐