均线滞后的时序自回归股市态势预测算法  

Time Series Autoregressive Stock Market Forecasting Algorithm Based on Moving Average Hysteresis

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科学与信息学院

出　　处：《郑州大学学报（理学版）》2018年第3期60-66,共7页Journal of Zhengzhou University:Natural Science Edition

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

摘　　要：针对艾略特波浪理论中的W形态,给出一种均线滞后性的量化方法,并提出融合均线滞后特征的时序自回归股市态势预测算法(DSMA).算法首先基于波浪理论提取W形态,给出W形态结点的量化表示形式;然后引入均线滞后性,并计算均线滞后程度;最后,利用贝叶斯网络表示融入均线滞后性的W形态结构关系,将各结点的局部关系代入AR模型中实现对股市态势的预测.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.Aiming at the W-shape and mean-hysteresis in Eliot′s wave theory,a time-dependent autoregressive stock market trend prediction algorithm(DSMA)was proposed.Firstly,according to the wave theory extracted W-shape,W-shape node quantification method was given.Then,the average hysteresis was introduced and the average hysteresis was calculated.Finally,the Bayesian network was used to represent the W-morphological relation into the moving average hysteresis,and the local relations of each node were substituted into the AR model to predict the stock market.The experimental results showed that the algorithm had higher prediction accuracy.

关 键 词：贝叶斯网络 结构关系 自回归预测模型 滞后性 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]