Clustering algorithm research and realization based on Local Gathering Features

Under the research and analysis on different types of clustering algorithms, focus on the limitation of the Jarvis-Patrick algorithm and other clustering algorithm based on SNN density, a new clustering algorithm is proposed in this paper, that is, Improved Clustering algorithm based on Local Gathering Features. The paper gives the definition of the Gathering Features during the procedure of the clustering, show the algorithm´s design and implementation method, and list the experimental data identification result. The new presented algorithm can deal with different types, dimensions, density and shape data collection problems, does not increase the time and space complexity, highlights the characteristics of Local Agglomerative Characteristics, improve the learning efficiency and the quality of data clustering.