Research On Clustering Algorithm Based On Swarm Intelligence

Data Mining is a superior area in the database and information technology, and is commonly considered as one of the key technology with wild developing perspective.Data Mining relates to machine learning,database,pattern recognition, AI and statistics technology etc.It is a new interdiscipline.Clustering is an important research field in Data Mining,and also an important method in data partition or data grouping.Cluster analysis as a module in the systerm of data mining can be used not only as a separate technique to discover the information about data distribution, but also as the preprocessing of other data mining operations.Therefore it is very meaningful to research how to improve the performance of clustering algorithms.Intelligent use of collective edge groups,in the absence of centralized control,a prerequisite for the overall model.In order to find solutions to complex problems with a new way of thinking,is a "characteristics of no wisdom main smart shown intelligent behavior by the main cooperation".Smart as the typical modes of the main groups,intelligent simulation of biological Ant colony algorithm optimization and simulation exercise habits of the PSO algorithm model is being widely academia concern.The main research of this paper is as follows:(1)Description of clustering analysis.The definition,data type,primary algoriths are reviewed.(2)Investigated the basic concept of swarm intelligence,and it's principle and main applications and so on.(3)The basic ant colony algorithm and LF model are introduced.It analyzes their advantages and disadvantages.Aiming at the drawbacks of LF model an improved LF algorithm is presented.The definitions of swarm similarity and probability transformation function are simplified.The parameters are adjusted adaptively.The experimental results show that the contradiction between convergence speed and clustering result is resolved in this improved algorithm.(4)The particle swarm optimization algorithm is introduced.The clustering algorithm based on particle swarm optimization algorithm with mutation is proposed.The problems of premature convergence and convergence speed in basic particle swarm optimization algorithm is resolved.The experimental results show that the algorithm not only avoids local optima, but also increases the convergence speed.