基于支持向量机的特征选择算法综述

特征选择是目前机器学习领域的研究热点之一,特征选择结果的好坏直接影响着分类器的分类精度和泛化性能.首先分析了特征选择算法的框架;其次分析了支持向量机用于特征选择的意义;然后对基于支持向量机的特征选择算法进行了分析和总结;最后从算法实用性角度出发,面向网络数据,探讨基于支持向量机的特征选择算法研究思路.

Feature Selection Algorithms Based on Support Vector Machine

Feature selection has been an important research field in machine learning, which influences directly the classification accuracy and generalization capability of a classifier. Firstly, the framework of feature selection algorithm is analyzed. The significance of using support vector machine （SVM） to select features is discussed. Then feature selection algorithms based on SVM are classified. Finally, from a practicality angle, using SVM to select features for network data is con- sidered, which could automatically get the relevant features just using the original network packet without a preprocessing step to get the original feature set.