基于实例学习和协同子集搜索的特征选择方法

摘要

特征子集搜索是数据挖掘分类任务中一个关键性的难题,常用的过滤器方法忽略了基因之间的相关性,此外,现有的解决方法并不是专门针对处理小样本数据,因此在特征选择方面表现出了不稳定性.为了解决上述问题,在实例学习的基础上提出了一种新型的混合封装过滤算法,并且提出了一种具有封装器评价体系的分类器算法——协同性子集搜索(CSS).选取几个高维小样本的癌症数据集作为数据来源,对提出的评价体系进行了实验测试,结果表明,该方法在准确性及稳定性方面较其他方法表现更好.%Feature subset selection is a key problem in such data mining classification tasks.In practice,the filter methods ignore the correlations between genes which are prevalent in gene expression data,additionally,existing methods are not specially conceived to handle the small sample size of the data which is one of the main causes of feature selection instability.In order to deal with these issues,a new hybrid,filter wrapper was proposed,and a cooperative subset search(CSS),was then researched with a classifier algorithm to represent an evaluation system of wrappers.The method was experimentally tested and compared with state-of-the-art algorithms based on several high-dimension allow sample size cancer data sets.Results show that the proposed approach outperforms other methods in terms of accuracy and stability of the selected subset.