Research And Application Of SN P System In Clustering Problem

With the development of information technology, data mining is attracting more and more attention of many experts and scholars, and it is being applied to the research and application to practical problems. Efficiency of data mining is mainly dependent on mining method. As a standalone tool or data mining algorithms(such as characteristics and classification) preprocessing step of other data mining, clustering has become a very active research topic in data mining area. The cluster analysis is an important branch in data mining area. Its purpose is dividing them into different classes based on different characteristics of different objects. At present, cluster analysis has been applied to many areas as well.Membrane computing is a new algorithm model that it regards biological cells as a prototype, and we study its structure and function, and follow the abstract and conversions. Membrane computing has unique characteristics,distributy, uncertainty, the largest parallel computing and membrane computing has many similarities to Turing machine. Their computing power is similar, but membrane computing is superior in some respects, especially in parallel and optimization issues. Spiking neural P system is an important branch of membrane computing, that the cell in it are used in neuronal cells in membrane systems, and their thoughts are from biological nervous system. Based on research status of Spiking neural P system, the current research results of SN P is theoretical studies which mostly focused on the validity and accuracy of the calculation, as well as a number of different Spiking neural P systems or variants based on different backgrounds Study. Achieving clustering mainly relies on traditional methods, in recent years,a number of biological calculation methods began to be used to solve the problem of the formation of a bio-cluster calculation method based on the new clustering ideas. The incorporation of membrane computing and clustering problem is a very important research direction. Under the guidance of this idea,this paper is put forward to solve the clustering problem with a special membrane system(Spiking neural P system), formation of a new clustering method based on SN P.In this paper, we mainly use parallelism and uncertain of Spiking neural P systems to achieve the system in Cluster. At first,we combine SN P with particle- clustering algorithm, to build a particle- clustering algorithm system based on SN P. Particle swarm algorithm use iterative randomness, that is,its full global optimization capability to find the optimized initial cluster centers, then iterate the k-means algorithm clustering in the case that the initial cluster centers are known. And by typical data of three different dimensions space in UCI dataset,we compare the standard PSO algorithm, PSO-KM algorithms in accuracy and time to relapse. SNPSOKM algorithm is designed with a more high accuracy and relatively little time to execute.This paper proposes a agglomerative hierarchical clustering model based on SN P with membrane computing features. It incorporates many features of membrane computing based on hierarchical clustering model, and then the traditional method has been improved and enhanced to reach the the computing power which original traditional methods can not achieve.Social network making-friends platform is an advanced business intelligence platform based on mass data mining, in order to help the user to select a good friend, and establish good relationships with each other. Social networking sites recommendation system recommends a friend for the user, and auto-complete a friend recommending process to meet the needs of users. The recommendation is based on: jointly owned friends, the number of co-owner friends and other hidden relationships. In this paper, the issue of a friend recommended is abstracted into hierarchical clustering problem, and therefore it is possible to use Spiking neural P system to achieve. It means being able to promote contact and communication between friends, and it has many important applications in the real world.