智能识别中的降维算法简述

An Introdution to Dimensionality Reduction Algorithms in Intelligent Recognition

近年来降维方法作为智能识别中关键的数据预处理技术得到了较为成功地运用。在总结比较经典的降维方法的基础上,分别从特征选择和特征提取两个方面来阐述各种方法所具有的特点和优势,并在一定程度上指出相应方法在目前所存在的问题和挑战。

Dimensionality reduction algorithms,as the key technologies of data preprocessing in intelligent recognition,have been used successfully recently.On the basis of the comparison of the classical dimensionality reduction algorithms,the characteristics and advantages of each algorithm are formulated from the perspectives of feature-selection and feature-extraction separately and the corresponding problems and challenges about each of them are pointed out to certain extent.