Research And Application Of Density Clustering Algorithm Based On Data Partition

Clustering analysis is the core technology of data mining with the process of dividing the physical or abstract collections into many clusters. Data objects are clustered with the principle of largest similarity in the same cluster and least similarity between different clusters. Clustering is the pretreatment of the data mining.Density-based clustering is qualified with fast, efficient in handling the problem of noise points and finding any shape. But in the data spaces which density is uneven distributed, it is hard to get the good results. This paper researched the density-based clustering algorithm DBSCAN and Partition DBSCAN in every detail. PDBSCAN used the mind of data partition and could get better quality of cluster in result on the data space whose density is uneven. But because the definition of he data partition of PDBSCAN is not precise. In this paper, we used the data partition algorithm redefine the data partition, precisely partitioning the density uneven data. Experiment proved that the result of the algorithm is better.Dalian National Economy Mobilization Potential Analysis System is based on the national economy investigation overall implementation of the program. Meanwhile, joining the characteristics of Dalian national economy mobilization potential, building the advanced and practical national economy mobilization potential database, for the purpose of providing the security of finishing the job which is organized and implemented National Economy Office efficiently and reliably, because that data mining analysis platform needs classifying some economic institutions by some indicators. So it needs the clustering algorithm to finish this job. In this paper, in order to solving the function of ensuring the most appropriate position of Center Of Wartime Casualties, we use the improved algorithm into the system and better the result of the clustering. And with this algorithm build the Center Of Wartime Casualties at at the most appropriate position. In this situation, we perfected the function of the data geography information decision making support.