Clustering Optimization Algorithm Research Based On Membrane System

Biological computing is a inter-discipline of computer science and bio-science.Membrane computing is an important branch of biological computing.The computational models investigated in membrane computing are called P systems.The P system has the advantages of the distribution,maximum parallelism and strong fault-tolerant.Clustering is an important part of data mining and is also an efficient way to analysis and process data.Various of clustering algorithms are put forward.Each of them has advantages and disadvantages and suit for different areas.With the increasing amount of data and the increasing complexity of the data types the traditional clustering methods can not meet the current clustering needs.The clustering optimal algorithm research has become a hot field.The study proposes efficient clustering optimization algorithms to solve complex clustering problem firstly.Secondly it designs the new membrane structure and membrane rules to compute the clustering algorithm and improve the effectiveness of the algorithm.Finally,the two membrane systems are applied in the field of data mining and image segmentation.The main work is as follows:In this paper,it first introduces the research background and the current research situation of membrane computing and clustering algorithms,the basic theories and methods of membrane computing and clustering.It also introduce the research object,structure,rule and calculation Framework of P system,the existing several categories of clustering algorithms and the current clustering algorithm with some improvements.What is more.We presentation the innovation and difficulties of this study.We put forward two core algorithms.The chapter 2 presents a new GD-K-medoids clustering algorithm and a new celllike membrane system to realize the method.The new algorithm combines the advantages of the grid clustering algorithms,the density clustering algorithms and the K-medoids algorithm.The P system with new structure and efficient rules has been designed to compute the novel algorithm.The experiment results proved the superiority of the present algorithm on the reducing the time complexity and improving the accuracy.The chapter 3 propose a MDE-K-means clustering algorithm based on the cell-membrane system.Different types of membrane rules are designed to help the operation of the clustering optimization algorithm.In this chapter we also compare the new algorithm with the old clustering algorithms to prove its better clustering results on the UCI Machine Learning Repository and the artificial data set.(3)The GD-K-medoids clustering algorithm is applied to the analysis of Macao tourists' consumption ability analysis.The official data and questionnaire are provided by Macao Tourism Bureau.The attributes related to consumption ability are extracted and After the data are cleaned,3577 data are remained.The data sets are divided into six categories.The evaluation and analysis for the experimental results help find more valuable knowledge for the Macao Tourism Bureau and help them to provide marketing strategy reference.(4)This study applies the MDE-K-means clustering algorithm to image segmentation to verify the effectiveness of the algorithm.The experimental results show that the proposed system is suitable for image segmentation.Compared with classical K-means algorithm and FCM clustering algorithm,the MDE-K-means has better segmentation effect,less iteration time,and fewer iterations.The last chapter of the study is the summary and outlook to sort out the whole content of the article.This study provides a new way for the clustering problems and also provides a new idea for the application of membrane computing.At the same time,the shortcoming of this study and the future research prospects also has been shown.