【作者】 张亮；

【导师】 邵必林；

【作者基本信息】 西安建筑科技大学 ， 管理科学与工程， 2009， 硕士

【摘要】 数据流是一种新型的数据模型,它的发展仅有二十几年的时间,数据流挖掘一直是数据挖掘和数据库领域中讨论的热点难点问题。本文在详细分析评述当前国内外数据流挖掘现状和挖掘算法的基础上,将更有效挖掘数据流中频繁项的MFI-TransSW算法首次应用在股票分析中,并详细介绍了MFI-TransSW算法的核心内容位序法的工作原理和挖掘频繁项的过程。文中将股票的相关性计算及模式依赖分析结合起来,能够更为准确地挖掘出股票之间的依赖关系。本文首次应用了MFI-TransSW算法,并提出了股票实时模式依赖挖掘模型,可以对具有模式依赖关系的股票进行挖掘和预测,即当一种股票出现特定模式时,另一种股票将出现什么样的模式。该文的研究成果可以对短期内的股票交易价格趋势进行合理的预测,满足用户对股市行情实时查询和分析的要求,为投资者的投资决策和股民选股提供一定的理论性指导,并为今后数据流挖掘算法的丰富和发展进行了有益的理论探索和实践尝试。更多还原

【Abstract】 Data stream is a new type of data model,and its development is only 20 years.Data stream mining has been a hot and difficult topic of in the field of data mining and databases.In this paper,a detailed analysis of the current review of the status based on at home and abroad and mining data stream mining algorithm the first time would make more efficient data stream mining of frequent itemsets in MFI-TransSW algorithm used in practice,and detailed information on MFI-TransSW algorithm rank the core content of the work of law principles and the process of mining frequent itemsets.In this article and share the relevance of model-dependent analysis of calculation and combination of the excavation can be a more accurate dependence between stocks.In this paper,the first application of the MFI-TransSW algorithm,and the stock mining model to rely on real-time mode,you can rely on the relationship with the stock model to carry out excavation and projections,that is,when a specific model stocks,and the other will be what the stock kind of model.The research results can be short-term trends in stock prices and reasonable projections,the stock market to meet the user’s query and analysis of real-time requirements for the investment decisions of investors and stock investors to provide theoretical guidance,and for the future the development of mining data streams to lay a certain foundation.更多还原