基于PCA和Rough Set在股票分类中的研究

Research of the Stock Classification Based on the PCA and Rough Set

为了提高股票分类的精确度,降低分类的复杂度,结合主成分分析和粗糙集理论对股票数据进行了处理。首先对股票数据进行预处理,然后利用主成分分析降低数据的维数,再利用粗糙集理论对降维后数据进行离散化和约简,并最终得到分类精度和分类规则。试验表明:算法取得了较好的分类精度和较少的分类规则数目,具有一定的可行性。

In order to achieve a better classification precision and a less classification complication of the stock, the stock data is disposed based on the PCA and rough set. First, the stock data is pre-treated, and the data dimension is reduced using the PCA. Then the data is dispersed and reduced using the rough set and gaining the classification precision and rules finally. The test shows that the algorithm can have a better classification precision and a less classification rule numbers and, so have feasibility in some degree.