一种基于层次聚类的流数据挖掘方法

流数据的特点在于数据流快速、有序地到达，并且数据海量，许多应用领域中生成的数据都可以归结为此类型．数据挖掘技术可以从海量的数据中发现有意义的知识模型，传统的数据挖掘算法通常是针对静态数据集，对流数据却无法有效地处理．文章试图从层次聚类角度处理流数据，并探讨了一种基于最小代价函数的层次聚类算法．

A Streaming Data Mining Method Based on Hierarchical Clustering

Streaming data are rapid,continuous,time-series and massive data elements.Many applications in different fields involve the data model.Data mining can discovery interesting knowledge patterns and have developed greatly in recent years.But,data mining algorithms which were put forword for traditional static data set can not mine interesting knowledge patterns effectively from streaming data model.Mining data streams needs special learning algorithms.The article proposes a hierarchical clustering mining algorithm on streaming data based on the minimum cost function.