基于特征聚类集成技术的组特征选择方法

Group feature selection based on feature clustering ensemble

提出了一种过滤冗余特征的算法框架,利用组特征选择算法去除冗余特征。在组构造阶段为了弥补单一聚类算法的不足,引入聚类集成的思想,先利用k-means方法通过多次聚类得到一个聚类集体,在集成阶段再利用层次聚类算法对聚类集体进行集成得到最终的结果。实验结果表明,这种算法框架能有效消除冗余特征,在保证算法稳定性的同时还能获得很好的性能。

The paper presents a filtering algorithm framework of the redundant features, which first take advantage of group feature selection algorithm to remove the redundant features. In order to make up the lack of a single clustering algorithm in the group formation, it puts forward the idea of clustering ensemble. First, it uses k-means clustering method to get multiple clustering. Then use hierarchical clu..;tering algorithm to integrate and obtain the final results Experimental results show that the algorithm framework can eliminate the redundant features effectively, at the same time ensure the stability without sacrificing the classification accuracy.