The Data Mining Based On Particle Swarm Optimization

As a new developed research field, data mining has developed rapidly in recent years and has widely applied in many fields. The stock market is not only an important component of the national economy, but also the significant way of people's investment. People press for an efficient tool which can analyze and deal with a lot of data of the stock market, so data mining has a good feature in stock market. Result from that many factors can impact on the latent rules of the data produced by the stock market and the dynamic changes of the these data, the research of data mining of the stock field has become one of the most popular researches of the world. Based on that the elementary feature determines on the stock in specific period and using the Particle Swarm Optimization (PSO), this paper builds a data mining system that can find out the items fit for investment in specific period.Firstly, we simply introduce the technology of data mining, including the basic models, regular algorithms and applied fields of the data mining. Based on that, we not only present the features and the applications of data mining in the stock field and point out the possibility of applying data mining in stock market, but also briefly introduce its current research situation.Secondly, we introduce the Particle Swarm algorithm and present the basic idea that PSO solves difficult problems by simple individual behaviors and studying each other. Comparing the PSO with other algorithms, we present the features, current research situation and applied fields of the PSO.Thirdly, using PSO and combining the theory of the data mining system with the features of the stock market, we bring forward a stock mining system based on PSO and particularly expatiate on the basic theories of the system, the define of the particle and the main problems of the pattern evolution.Finally, based on researches above, we implement the data mining system of the stock with aggregation database and object oriented technology. Then, we descript in detail not only the main models of the system but also the design and the implement of the data structure, mine the related data and verify the rationality and possibility ofthe system.