Research On Clustering Algorithm Of Multidimensional Spatial Data Complexity Features

With the rapid development of Internet applications and the rapid development of the Internet of Things,the types of data have become more and more complex,and the number of data has become more and more.People want to use these diverse data to better serve the public,so they have carried out research on these data.Because of the complexity of data,multidimensional data and obstacle constraint data are also derived.This paper mainly focuses on multi-dimensional data and obstacle constraint data.Firstly,a new multidimensional data clustering algorithm based on CLIQUE algorithm is proposed to solve the problem of manually selecting grid length and density thresholds and grid boundary hard partition when clustering multidimensional data.Then,a multidimensional data clustering algorithm is proposed to deal with obstacle constrained data.Aiming at the problem that the classical CLIQUE algorithm in multidimensional clustering algorithm randomly selects the density threshold and is sensitive to the density threshold,and the problem that the random selection of grid length causes hard grid boundary division,a multidimensional data clustering algorithm based on adaptive grid boundary division(AGB-Clus)is proposed.First,the algorithm uses the density threshold calculation formula to calculate the density threshold of each dimension.Secondly,according to the number of closely connected unit grids in each dimension,the subspaces that meet the clustering requirements are selected,and the complexity of the algorithm is reduced by filtering the subspaces.Finally,the adaptive grid boundary is obtained by calculating the initial grid length,so as to determine the cluster and complete the data clustering.The experimental results show that the proposed algorithm is compared with the experimental comparison algorithm.The algorithm in this paper has good clustering effect and accuracy.In order to effectively solve the problem of data clustering with obstacle constraints,a multi-dimensional data clustering algorithm based on obstacle constraints(OBPrim DPA)is proposed on the basis of adaptive grid boundary division algorithm.First,for data with obstacle constraints,the grid is divided by using the density threshold calculation formula,and the grid whose density is higher than the density threshold is divided into the dense cell grid set;The grid with grid density equal to 0 is divided into empty element grids;Adjacent dense cell grid threshold is used to judge the remaining undivided grid.If the number of dense cell grids in the adjacent cell grid of the current grid is greater than or equal to the adjacent dense cell grid threshold,it is directly placed in the dense cell grid set.If it is less than the adjacent dense cell grid threshold,the adaptive boundary division algorithm is used to divide the remaining divided grid,Get multiple densely connected units,and then calculate the distance between data points for each densely connected unit to judge the impact of obstacles on data points,and finally get the clustering result.The experimental results show that the proposed algorithm is compared with the experimental comparison algorithm.The algorithm in this paper has good clustering effect and accuracy when clustering data with obstacle constraints.