论文部分内容阅读
为了全面评价特征子集的好坏,提高特征子集作为最佳子集的可靠性,以及更快找到最佳子集,提出了一种用于特征选择的多准则融合差分遗传算法。引入多个评价准则对特征子集进行评价,并对遗传算法的选择算子进行改进,有利于选出适应度高且具有重要特征的个体;同时,引入差分策略改进变异算子,提高种群多样性和算法搜索能力;最后通过仿真实验和滚动轴承实例验证了该方法的有效性。
In order to comprehensively evaluate the quality of feature subsets, improve the reliability of feature subsets as the best subsets, and find the best subsets faster, a multi-criteria fusion genetic algorithm for feature selection is proposed. The introduction of multiple evaluation criteria to evaluate the feature subset and improve the selection operator of genetic algorithm is conducive to the selection of individuals with high fitness and important features; the same time, the introduction of differential strategy to improve the mutation operator to improve population diversity Sex and algorithm search ability. Finally, the effectiveness of this method is verified by simulation experiments and rolling bearing examples.