基于文化基因算法和最小二乘支持向量机的安全数据特征处理方法

Feature Processing Approach Based on MA-LSSVM in Safety Data

随着各类生物智能演化算法的日益成熟,基于演化技术及其混合算法的特征选择方法不断涌现。针对高维小样本安全数据的特征选择问题,将文化基因算法(Memetic Algorithm,MA)与最小二乘支持向量机(Least Squares Support Vector Machine,LS-SVM)进行结合,设计了一种封装式(Wrapper)特征选择方法(MA-LSSVM)。该方法利用最小二乘支持向量机易于求解的特点来构造分类器,以分类的准确率作为文化基因算法寻优过程中适应度函数的主要成分。实验表明,MA-LSSVM可以较高效地、稳定地获取对分类贡献较大的特征,降低了数据维度,提高了分类效率。

With all kinds of biological intelligent evolutionary algorithms become increasingly mature, feature selection methods based on the evolutionary technology and its hybrid algorithm are emerging. According to the feature selection problem of the high dimensional small sample safety data, this paper combined memetic algorithm （MA） and least squares support vector machines （IAS-SVM） to design a kind of wrapper feature selection method （MA-LSSVM）. The proposed method utilizes the specialty of least squares support vector machine being easy to search optimal solution to construct classifier, then regards classification accuracy as the main component of memetic algorithm fitness function in the optimization process. The experimental results demonstrate that MA-LSSVM can be more efficient and stable to ob- tain features larger contribution to the classification precision, and can reduce the data dimension and improve the classi- fication efficiency.