关联规则在股票分析中的应用

应用Apriori算法的思想,采用变支持度和变置信度进行股票数据的一维和多维关系的 挖掘。不同于传统的以股票代码作为数据预处理的结果,本文采用股票的交易日期作为预处理的结 果,把交易日期所在的列号作为算法的直接处理对象,把交易日期作为其间接的处理对象。这样可以 方便快捷地挖掘出用户感兴趣的规则,同时也避免了在进行多维数据挖掘时需进行再处理的麻烦。 有代表性规则的验证证实了挖掘结果的正确性。

Application of association rules in stock analysis

The Apriori arithmetic adopted the changing support and confidence was applied to mine one-dimension and multi-dimension stock information. It was familiar to take the stocks' codes as the results of data processing, but the stocks' transaction time was selected as the results of data processing aiming at the convenience and celerity during the concrete mining. It took the tier of transaction time as the direct object of arithmetic of data mining and took the transaction time as the indirect object of arithmetic of data mining. After such data processing, it was effective not only to mine the interesting rules, but also to avoid the trouble that the results of the data processing need to be reprocessed during the mining of multi-dimension information. The mining results have been proved to be right by the validation of the representational rules.