摘要
随着我国股市的逐步规范和完善,投资者愈加重视投资对象的选择。针对股票分类的特点,选取对上市公司股票走势有重要影响的7项主要财务指标,运用聚类分析和支持向量机相结合的方法对80家上市公司的股票进行分类。为了降低分类实验的复杂程度,在分类实验中采用因子分析法将原来的7项财务指标用3个综合指标概括。实验结果表明,这种方法大大降低了股票数据的维数,有很高的分类正确率,分类达到了让人满意的效果,证明这种分类方法是可行的。
As China's stock market gradually standardized and improved, investors paid more attention to the selection of investment targets. Based on the characteristics of the stock classification, selected the data that greatly influenced the stock development trend of listed companies. 7 financial ratios from 80 stocks of listed companies had been studied by means of chimer analysis and support vector machines. Aiming at reducing the data complexity of the classification experiment, 7 primary financial ratios were summarized by three comprehensive indexes in classification experiment in the way of factor analysis. According to the experimental results, the method introduced in this paper reduced the dimensions of stock data greatly, and it had high classifieation accuracy. The satisfactory results of classification proved its feasibility.
出处
《计算机技术与发展》
2009年第6期229-231,共3页
Computer Technology and Development
基金
辽宁省高等学校科研项目(2008343)
关键词
因子分析
聚类分析
支持向量机
股票研究
分类
factor analysis
cluster analysis
support vector machines(SVM)
stock research
classification