融合单类F-score和遗传算法的微生物特征选择方法  

Microbial feature selection method combining one class F-score and genetic algorithm

作　　者：卢福梅 温柳英[1] LU Fu-mei;WEN Liu-ying(School of Computer Science,Southwest Petroleum University,Chengdu 610500,China)

机构地区：[1]西南石油大学计算机科学学院,成都610500

出　　处：《信息技术》2024年第11期125-131,共7页Information Technology

基　　金：中央引导地方科技发展专项项目(2021ZYD0003);西南石油大学启航计划(2018QHR007)。

摘　　要：微生物数据由于其维度高和类别不平衡特点,在传统机器学习分类算法中不能得到理想的分类效果。传统特征选择算法可以达到降维的效果,但对于类别不平衡问题显得有点力不从心。因此,文中提出了一种融合单类F-score和遗传算法的特征选择方法。首先,利用单类F-score操作生成遗传操作中的初始种群;其次,利用SVM分类模型的AUC值作为遗传操作中个体的适应度值;再次,结合遗传操作来更新种群;最终,得到最优特征子集。实验在五个微生物数据集上进行,与四种特征选择算法进行对比,结果表明,文中所提方法在一定程度上优于其他方法。Due to its high dimensionality and imbalanced categories,microbial data cannot achieve ideal classification results in traditional machine learning classification algorithms.The traditional feature selection algorithm can achieve the effect of dimension reduction,but does not work in terms of category imbalance.Therefore,this paper proposes a feature selection method that combines single class F-score and genetic algorithm.Firstly,the initial population in the genetic operation is generated using a single class F-score operation.Secondly,the AUC value of the SVM classification model is used as the fitness value of individuals in genetic operations.Thirdly,genetic operations are combined to update the population.Finally,the optimal subset of features is obtained.The experiment is carried out on five microbial datasets and compared with four Feature selection algorithms.The results show that the method proposed in this paper is superior to other methods to some extent.

关 键 词：F-score 高维 不平衡 遗传算法 特征选择 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]