The Association Rule Mining Algorithms And Application Based On Swarm Intelligence

Swarm intelligence is a new method of optimization witch groups in the past 10 years on the basis of f evolution in coordination developed. It is complicated to be a single individual to complete a complex task to a large number of simple individual components to complete the task, and the latter is more robustness, flexibility and economic advantage. 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. Due to its simple concept, easy to be realized in the short run was quickly recognized by the international field of evolutionary studies.Association Rules Mining found a large amount of data associated with or related to the focus of the link between interesting. It is an important issue in data mining, then has been extensively studied by the filed in recent years. Rules research helps to find the link among different items late transactions database ,find out customer buying behavior mode, the results of the analysis can be applied to merchandise shelf layout, survival mode, as well as to classify the users under the purchase mode. Agrawal was first proposed in 1993 equivalent to Mining Association Rules about sets of items in customers, transaction database. After then many researchers made a lot of research on the issue of mining association rules .In this paper give a more systematic analysis and research against swarm intelligence especially PSO algorithm, combining association rules algorithm proposed some improved algorithm , including the following :(1) Swarm Intelligence Technology outline. The basic concept of swarm intelligence, classification, the major functions, key technology and typical application and so on.(2) Mode of swarm intelligence. Introduced two Common modes of the swarm intelligence, made the comparison through specific examples of contemporary international advanced swarm intelligent algorithms, and reached a conclusion.(3) Summary of the Mining Association Rules. This article made the analysis on representative association rule mining algorithms .proposed typical requirements of mining association rules and based on these requirements made the comparisons on the association rule mining algorithm used, in order to facilitate more easily. Choose a more rapid mining algorithm applied to specific problems.(4) The sampling based on PSO algorithm applies in the application of association rules. Propose an improved algorithm for mining association rules. While in theory the traditional one-time sampling algorithm to ensure the precision of the results, when used in association founds of many types of commodities and very large transactions, the calculation time is substantial, and even is not limited in the year and hard in line analytical application, which would restrict the use of Apriori algorithm. Sequential sampling theory algorithm design can be upgraded ASAR. APASAR algorithm can produce frequent set effectively, and the needs of the samples less than one-time sampling algorithm. However, the sample size wills still theoretical high when performance. This paper presents a kind of swarm intelligent technology Particle Swarm Optimization (PSO) algorithm combined with the algorithm Apriori, called PSASAR algorithm. Using the intelligent technology to accelerate the process, to increase intelligence, when database scanning, the use of intelligent technology group, replace the entire database scan, enhance efficiency, scanning effective than random sampling.