论文部分内容阅读
鉴于支持向量机特征选择和参数优化对其分类准确率有重大的影响,将支持向量机渐近性能融入遗传算法并生成特征染色体,从而将遗传算法的搜索导向超参数空间中的最佳化误差直线.在此基础上,提出一种新的基于带特征染色体遗传算法的方法,同时进行支持向量机特征选择和参数优化.在与网格搜索、不带特征染色体遗传算法和其他方法的比较中,所提出的方法具有较高的准确率、更小的特征子集和更少的处理时间.
In view of the fact that SVM feature selection and parameter optimization have a significant impact on the classification accuracy, the asymptotic performance of SVM is incorporated into genetic algorithm and the feature chromosome is generated, which leads the search of GA to the optimization error in hyperparametric space Based on this, a new method based on genetic algorithm with feature chromosome is proposed, and the feature selection and parameter optimization of support vector machine are carried out at the same time.Comparing with grid search, genetic algorithm without feature chromosome and other methods , The proposed method has higher accuracy, smaller feature subset and less processing time.