摘要
典型相关分析(CCA)在执行分类任务时主要存在如下不足:1)尽管分类时的输入是组合特征,但CCA仅优化组合特征的各组成部分,并未直接优化组合特征本身;2)尽管面对的是分类任务,然而CCA根本无法利用样本的类信息.为弥补CCA的上述不足,文中提出一种监督型降维方法——增强组合特征判别性的典型相关分析(CECCA).CECCA在CCA基础上,通过结合组合特征的判别分析,实现对组合特征相关性与判别性的联合优化,使所抽取特征更适合分类.在人工数据集、多特征手写体数据集和人脸数据集上的实验结果验证该方法的有效性.
Canonical Correlation Analysis （CCA） has following two deficiencies m pertormlng classmcanon task： CCA can not directly optimize them but their components, though combined features are the input of the classifier; CCA can not utilize any class information at all, though facing classification task. To overcome these deficiencies, a supervised dimension reduction method named combined-feature- discriminability enhanced canonical correlation analysis （CECCA） is proposed. CECCA is developed through incorporating discriminant analysis of combined features into CCA. Consequently, it optimizes the combined feature correlation and discriminability simultaneously and thus makes the extracted features more suitable for classification. The experimental results on artificial dataset, multiple feature database and facial databases show that the proposed method is effective.
出处
《模式识别与人工智能》
EI
CSCD
北大核心
2012年第2期285-291,共7页
Pattern Recognition and Artificial Intelligence
基金
国家自然科学基金(No.61170151)
江苏省自然科学基金(No.BK2011728)资助项目
关键词
典型相关分析(CCA)
分类
降维
组合特征
信息融合
Canonical Correlation Analysis （ CCA）, Classification, Dimension Reduction, CombinedFeatures, Information Fusion
