Research And Application Of Partition Clustering Based On Organizational P System

With the development of information society,the demand for the quality of largescale data processing is getting higher and higher.Great progress has been made in the development of both hardware and software.The hardware has biological computer,quantum computer,,photon computer.Membrane computing is an important research direction in biological computing,and its distributed maximal parallelism compute method has attracted the attention of many scholars.In the information society,the processing of a large amount of information data becomes inevitable.Clustering,as an important algorithm of data mining and knowledge discovery,is an important way of data processing.After many years of research,a large number of clustering algorithms have been invented.These clustering algorithms have their own advantages and disadvantages,and different clustering algorithms are suitable for different data types or special scenes.Researchers are working on more efficient and versatile algorithms.In this paper,the tissue-like p system and the partition clustering algorithm are studied.A general tissue-like p system is put forward first,then the clustering algorithm is optimized.The calculation process of the algorithm is simulated by p system.The comparison experiment shows that the improved algorithm has better running efficiency.The main tasks are as follows:The research background and significance of this paper are introduced.The present situation and development trend of membrane computing calculation and cluster analysis at home and abroad are summarized in detail.The research status and development trend of the combination of membrane computing and cluster analysis are also briefly introduced.Then an overview of the time controlled tissue-like P system mentioned in this paper is introduced,and the basic theory of partition clustering analysis is introduced.In this paper,we proposed an priority variable tissue p system.The rule priority of the traditional p system is fixed.In general,the rule whose priority is forward executed first,and the rule whose priority is in the rear executed later.However,when the clustering algorithm is implemented,some cyclic operations are often not well implemented.The p system we propose after executed the rule whose priority ahead once should wait the rule whose priority backward executed,even if it can be executed the seconed time,and the like.Such a provision can easily realize circular operation.Secondly,we have improved the ranked k-medoids algorithm,the maximum distance method is introduced,and the initial clustering center is selected by the maximum distance method,which avoids the algorithm falling into the local optimum and reduces the complexity of the algorithm and improves the stability of the algorithm,before and after the improvement,the algorithm is only suitable for the data structure based on Gaussian distribution.PAM algorithm is robust,but its time complexity is high,so it is not suitable for large scale data clustering.This paper combines PAM algorithm with KNN algorithm,Reduce the number of loss comparisons from all points in the cluster to our predetermined number,these points are the nearest m points to the center point,so the time complexity of the original PAM algorithm is greatly reduced.It is more suitable for large data clustering.For these two algorithms,we use the time-controlled organization p system based on the object control flow to design the rules,so as to complete the whole calculation process,object control flow plays an important role in computing the distance between data points and the time that rules to be executed.In the membrane system,it is parallelism between cells and rules,which greatly improves the efficiency of the algorithm.We apply the KNN-PAM algorithm to the division of high-tech enterprises.Different high-tech enterprises have different levels of innovation,and we classify hightech enterprises with the same or similar levels of innovation.Then make different policies for different high-tech enterprises.Provide different development ideas for different high-tech enterprises.Similarly,we will apply this system to carry travel net official micro user division.