Abstract: On basis of analyzing the advantages and weaknesses of the current clustering algorithm of data streams, this paper introduces a new density-based clustering algorithm in evolving data streams——SDStream, which can analyze and deal with large-scale evolving data stream. Its performance is tested by using both real datasets and synthetic datasets. Experimental results show this algorithm has better perpformance of applicability, effectiveness and extension and achieves a higher quality of clustering.

Key words: data mining, evolving data stream, clustering, sliding window

摘要： 分析当前数据流聚类算法的优点及不足，提出一种新的进化数据流中基于密度的聚类算法——Sdstream算法，该算法能够分析并处理大规模进化数据流，利用真实数据集和仿真数据集对其进行性能测试，实验结果表明，该算法具有良好的适用性、有效性和可扩展性，能够取得较高的聚类效果。

关键词: 数据挖掘, 进化数据流, 聚类, 滑动窗口

CLC Number: 

• TP311