基于遗传算法学习聚类算法的中心个数

无导师聚类算法的目标是将一个数据集划分为若干个类,使得类内相似性尽可能大且类间相似性尽可能小。聚类过程中对数据集合分割成多少个类是一个很难确定的问题,目前还没有较好的解决方法。文章使用遗传算法对无导师聚类K-均值(K-means)算法中中心个数K值进行学习,实现了使用遗传算法进行聚类中心个数的确定,旨在提供一种选择中心参数个数的方法。通过对UCI机器学习数据库中的7个数据库进行实验,证实此方法是比较有效的。

Learning the Center Number of Clustering Algorithms Using Genetic Algorithms

Unsupervised clustering algorithms partitions a data set into several groups such that the similarity within a group is larger than that among groups.It is difficult to decide the number of groups in the process of clustering,and no better solution has been put forward until now.In order to offer an method for parameter selection of partition num-ber,this paper proposes to learn the parameter k in unsupervised clustering K-means using genetic algorithms ,such that the number of clustering partitions can be achieved before clustering.Seven databases,which are selected from UCI repository of machine learning databases,are used to prove the validity of this algorithm.