Data stream clustering algorithm based on grid and density

According to the characteristics of the data stream,a new clustering algorithm GTCS which combined the approach based on density and grid was presented.By means of the model of double-layer construction,the method set the key of densities of the data grids automatically in online layer.The offline layer using the data gravity for the center,a maximum of subgrid is built.It makes the dense regions of the candidate grids into dense grid.It uses the minimum spanning tree clustering algorithm to get the clustering results and improve the clustering affect.