摘要
对股票的预测研究一直是社会的热点问题,而一个合适的模型对于股票的研究至关重要。选用加权马尔可夫链模型,对该模型的状态划分区间做了优化改进,使得该模型能更好地应用于股票的预测研究。通过对招商港口股票的收益率建立加权马尔可夫链模型,并与马尔可夫链模型的相应预测结果进行对比,在状态预测上,加权马尔可夫链的准确率提高了一倍,在平均股价绝对误差上降低了2.02%。此外,为说明模型的泛化性,增加了10只股票3个日期共30个数据预测股价,其与实际股价在每只股票每个日期下的平均绝对误差为1.67%,相比马尔可夫链模型的相应结果降低了0.51%,这表明加权马尔可夫链模型对于多种股票的收益率和股价的预测均具有合理性。
The research of stock prediction has been a hot issue in the society,and a suitable model is very important for the re⁃search of stock.This paper selects the weighted Markov chain in which the state partition interval of the model is optimized to make the model better applied in forecasting stock.By establishing a weighted Markov chain model for the return rate of China Merchants Port stock,and comparing the corresponding prediction results with those of the Markov chain model,the accuracy of the weighted Markov chain is doubled in the state prediction,and the absolute error of the average stock price is reduced by 2.02%.In addition,in order to illustrate the generalization of the model,we add 30 data of 10 stocks in 3 days to predict the stock price,and the aver⁃age absolute error between the actual stock price and the predict stock price in each day is 1.67%,which is 0.51%lower than the corresponding results of the Markov chain.This shows that the weighted Markov chain model is reasonable for the prediction of the rate of return and stock price of various stocks.
作者
陈梓海
黄香香
CHEN Zihai;HUANG Xiangxiang(School of Computer Science and Technology,Dongguan University of Technology,Dongguan 523808,China)
出处
《东莞理工学院学报》
2023年第3期1-8,共8页
Journal of Dongguan University of Technology
基金
国家自然科学基金(11801073)
广东省自然科学基金(2017A030310598)。
关键词
加权马尔可夫链
转移概率
收益率的预测
股价的预测
weighted Markov chain
transition probability
prediction of the rate of return
prediction of the stock price
