一种结构成熟度的时序自回归股市预测算法  

A structure maturity timing autoregressive algorithm for stock market prediction

作　　者：姚宏亮[1] 艾刘可 王浩[1] 李俊照[1] YAO Hongliang;AI Liuke;WANG Hao;LI Junzhao(School of Computer and Information,Hefei University of Technology,Hefei 230009,China)

机构地区：[1]合肥工业大学计算机与信息学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2018年第11期1484-1490,共7页Journal of Hefei University of Technology：Natural Science

基　　金：国家自然科学基金资助项目(61175051;61175033);国家重点基础研究发展计划(973计划)资助项目(2013CB329604)

摘　　要：由于现有的时序预测方法仅利用局部的组合特征,信息融合低,预测效果不佳。文章基于艾略特波浪理论中的W形态,通过结构建模,提出一种结构成熟度信息的时序自回归股市预测算法。首先提取股市中的W形态,通过量价波动对W形态均线走势影响分析,引入W形态结点成熟度,并且给出了价格的波动对均线走势影响的证明,利用贝叶斯网络表示W形态中结点成熟度之间的结构关系,实现对波段信息的有效融合;然后利用非对称信息熵计算结点间的关系强度,生成基于AR的结构成熟度模型;最后利用该模型进行股市态势预测。在实际数据上进行了算法比较分析,实验结果表明算法具有更高预测精度。As the existing timing prediction methods only use the local combination characteristics, the degree of information fusion is low and the effect of prediction is not good. Based on W shape in Elliott wave theory, the structure maturity algorithm is proposed through modeling the system structure, which is a timing autoregressive algorithm for stock market prediction. The W shape of stock market is first extracted, and the moving averages trend is analyzed by the volatility between the volume and price returns. The concept of node maturity is introduced, and the impact of price fluctuation on the moving averages trend is proved. In addition,Bayesian network is used to denote the structural relationship among nodes＇ maturity in W shape, and the effective integration of band information is achieved. Next, the structure maturity model based on AR is created by using the strength of ties among nodes that can be calculated with asymmetric information entropy. Final ly, the trend of stock market is forecast by using this model. The comparison of the algorithms based on real data is also carried out, and the simulation results show that the proposed algorithm has higher prediction accuracy.

关 键 词：贝叶斯网络 结构关系 自回归预测模型 W形态 波浪理论 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]