Research On The Improved Particle Swarm Optimization Based On The P Systems And The Application In The Clustering Analysis

Membrane computing is also called membrane systems or P systems, which are distributed and parallel computing model. A large amount of investigates had shown that, many classes of P systems are able of simulating Turing machines, hence they are computationally complete. Given the powerful parallel computing ability,membrane systems is likely to break through the limitation of the Turing machine, or instead of it. Membrane system has become a hot research field,and it will develop rapidly in various fields.Clustering is the process of grouping a series of data objects into multiple groups or clusters so that intra-cluster data are similar and inter-cluster data are dissimilar. It is widely used in many fields, such as machine learning, image pattern recognition,market analysis and so on. Particle swarm optimization(PSO) is a stochastic optimal evolutionary algorithm, it simulates the social behavior of birds flocking or fish schooling in order to achieve a self-evolving system. As a population intelligence algorithm, the Particle Swarm Optimization is widely used in optimization domain Optimization algorithm has been used successfully in many fields because of its features of being easily implementation and having fewer parameters, such as data clustering, pattern recognizing, neural network training and so on. This algorithm is easily trapped into the local optimum and has premature convergence phenomenon, in order to overcame the drawbacks and improve this situation, many improvements have done to the algorithm.This paper makes some appropriate improvements based on particle swarm optimization algorithm, by means of avoiding the worst idea for optimization. This idea does help in improving the global search capacity and refraining from trapping into the local optimum. The mutation mechanism could improve the diversity of the swarm, by introducing the genetic mechanisms of genetic algorithm, the particle swarm optimization is easy to jump from the local optimum. In the standard particle swarm optimization, the change of location is straightly, therefore, we insert simulated annealing into particle swarm optimization algorithm to make probability rules for location changing, this measure could raise the capacity of global search. In this paper,we tried to modify the particle swarm optimization based on this algorithms, and then combine the improved particle swarm optimization algorithm with the membrane system, instead the membrane rules of evolutionary algorithm, communication the optimum by communication rules, in other words, a membrane clustering algorithm is proposed in order to realize partition clustering. In response to the problems of the existing clustering algorithms, we try to make some modifications on the existing membrane evolutionary method and help to improve the performance of the algorithm.Customer relationship management (CRM) is a kind of management model which takes the customer as the center, dominated by enterprises and external communications, gives priority to with the front-end business application. Customer segmentation is the basis of customer relationship management, it divided a large customer base or consumer groups into several segments, in these groups, the customers belong to the same segment have similar consumption characteristics ,and have different consumption features from that of different segmentation.Segmenting customers by the method of data mining, we can help enterprises understand the characteristics of customers' behavior, find new business opportunities,reduce the marketing costs for both the customer and better meet the demand of consumers better. This paper studies and realizes the customer segmentation of customers in the tourism industry with the improved K-means algorithm. The customers of different groups are divided into different groups. The customers in the same group are similar to the requirements of tourism, while the customers of different groups have different tourism requirements. The enterprises could make different arrangements for different types of tourists,these measures could reduce costs while achieving maximum customer satisfaction, improve the company's efficiency and competitiveness, help enterprises stand in an impregnable position in in the increasingly fierce market competition.