A new supervised learning hierarchy clustering classification method

Cluster refers to the classic unsupervised learning method, whose mission is to group the samples as similar as possible within the cluster and between-cluster dierences as large as possible. The cluster is mainly divided hierarchical clustering, model-based clustering based clustering and density grid-based clustering and other types [2]. Due to the lack of sample class labels these teachers supervise signal, the parameters such as the number of clusters should be pre-specified in the cluster when the sample density was not significantly dierent, with low eciency and unknown actual classes of clusters.