摘要
研究股票数据的特点,根据传统的SVM算法原理,提出一种更新的动态预测模型(DMDI)对股市的大盘和个股的指数分别进行中短期预测。大量的数值实验结果表明,改进的支持向量机算法比单纯直接用支持向量机对股票指数的预测更为有效,具有明显的优越性。
The article studys the characteristics of the stock data and presents a dynamic model which based on the traditional SVM arithmetic. It applies the new algorithm in the prediction of the stock index. The results show that the algorithm is effective and correct. It has an obvious advantage of the traditional methods.
出处
《重庆科技学院学报（自然科学版）》
CAS
2008年第2期59-62,共4页
Journal of Chongqing University of Science and Technology：Natural Sciences Edition
关键词
支持向量机
股票指数
核函数
动态选择
support vector machine
stock index
kernel function
dynamic selection
