Cyber physical systems have grown exponentially and have been attracting a lot of attention over the last few years. To retrieve and\nmine the useful information from massive amounts of sensor data streams with spatial, temporal, and other multidimensional\ninformation has become an active research area. Moreover, recent research has shown that clusters of streams change with a\ncomprehensive spatial-temporal viewpoint in real applications. In this paper, we propose a spatial-temporal clustering algorithm\n(STClu) based on nonnegativematrix trifactorization by utilizing time-series observational data streams and geospatial relationship\nfor clusteringmultiple sensor data streams. Instead of directly clusteringmultiple data streams periodically, STClu incorporates the\nspatial relationship between two sensors in proximity and integrates the historical information into consideration. Furthermore,\nwe develop an iterative updating optimization algorithm STClu. The effectiveness and efficiency of the algorithm STClu are both\ndemonstrated in experiments on real and synthetic data sets. The results show that the proposed STClu algorithm outperforms\nexisting methods for clustering sensor data streams.
Loading....