摘要
提出了一种基于EMD(经验模态分解)与RVM(相关向量机)的股指期货价格预测方法。首先将目标价格序列通过EMD技术进行分解,然后对分解后的分量进行重组得到三个新序列,通过分析这三个新序列的特点,构造不同的RVM模型对每个新序列分别进行预测,最后将三个新序列的预测结果通过RVM组合得到最终预测值。实验结果表明,该方法能取得良好的预测效果。
This paper proposes a price forecasting method of Stock Index Futures based on EMD and RVM. First, the target price series can be decomposed by using the EMD algorithm. Second, the components above are restructured into three new sequences. By analyzing the characteristics of the three new sequences, different RVM models are constructed to predict values of each sequence. The final predictive value can be obtained by combining the predicted values of three new sequences with RVM. Experimental results show that the method can predict good results.
出处
《武汉工业学院学报》
CAS
2013年第2期115-118,共4页
Journal of Wuhan Polytechnic University
