Sequential Pattern Mining Optimization Algorithm And Its Application In Stock Investment

Data mining is usually combined with statistics, machine learning,information retrieval, expert systems and pattern recognition, etc., used for research and other aspects of database and artificial intelligence, and gained considerable progress. Association rules analysis is found in connection rule is different between transactions in the transaction database and association rules found to help people in the operation of the market, decision support and business management, to provide effective information, association rules in practical applications coverage is very wide, and occupies a more important role in data mining. As China’s market economy continues to develop and the growing size of the stock market, the role of stock markets in China’s growing economic life. Due to the size of the stock market, the data are more numerous, the law of stock price performance is not obvious, but hidden in thousands of transactions database. Therefore, how these transaction history data isfully utilized, and the use of mining association rules to find hidden in the data behind the rules, which the people on the stock market forecasts and investment to provide favorable value, association rule mining algorithm application Research and Exploration in the field has become meaningful.This article describes the data in the mining association rules and the basic theory and research status of background knowledge, combined with mining stocks timing characteristics and needs, it leads to the sequence mode with the rules of the time factor of the definition of mining. Classical algorithm Apriori algorithm for association rules bottlenecks and lack of sequential pattern mining traditional sliding window technique is proposed to optimize the sequence mode improved algorithm, not only in the algorithm improves the efficiency and reduces the space burden, and the use of mode confinement and the interest of the constraints to filter the user is interested in and meaningful rules, such a rule for investment decisions more valuable. Finally, the optimization algorithm is applied to sequential pattern stock 2013-2015 two-year period of mining, and before and after the time of the stock plate linkage rules are compared with the results of the rules prevailing market policies combined analysis, interpretation of the rules effectiveness and optimization algorithm, but also has certain reference value for the investment later analysis.