HPDBSCAN:高效的不确定数据处理算法

HPDBSCAN: Efficient clustering algorithm for processing uncertain data

为了解决无线传感器网络中的不确定数据,提出了一种无线传感器网络不确定数据高效处理算法。根据不确定性数据的概率密度分布进行概率聚类,并利用Hilbert编码技术将多维数据映射到一维数据空间,通过基于Hilbert-R树索引的不确定性数据HPDBSCAN算法对不确定性数据进行聚类。实验结果表明,HPDBSCAN算法预处理效果较好,比其它聚类算法更适合不确定性数据的聚类。

To solve the uncertain data in wireless sensor networks, an efficient algorithm for processing the wireless sensor net work uncertain data is presented. According to the data of the uncertainty probability density to process distribution probability clustering, and use Hilbert encoding technology mapping the multidimensional data to a 1D data space, through the uncertainty of the data HPDBSCAN algorithm based on HilbertR tree index for the uncertainty data clustering. The experimental results show that HPDBSCAN algorithm pretreatment effect is better than other clustering algorithm and it is suitable for more uncer tainty data clustering.