CSA  >> Vol. 7 No. 7 (July 2017)

    基于ARIMA模型与BP神经网络的三七价格预测
    Price Estimation of Pseudo-Ginseng Based on ARIMA Model and BP Neural Network

  • 全文下载: PDF(3087KB) HTML   XML   PP.696-710   DOI: 10.12677/CSA.2017.77081  
  • 下载量: 145  浏览量: 205   科研立项经费支持

作者:  

刘建中:天津市国资委,天津;
李慧超:南开大学计算机与控制工程学院,天津

关键词:
三七价格预测ARIMA模型BP神经网络Pseudo-Ginseng Price Forecast ARIMA Model BP Neural Network

摘要:

本文分别选取了ARIMA、BP神经网络方法对三七价格进行分析预测。首先,采用传统时间序列分析模型——自回归移动平均模型对三七价格进行预测,发现其仅能提取价格变化曲线中的线性特征,对于离群点拟合效果不佳。BP神经网络适用于非线性可分问题求解,结合三七价格数据的波动性和非线性的特征,使用BP神经网络对三七价格进行分析预测,并利用相空间重构方法确定网络输入维数,加快网络结构确定。本文以云南文山地区2010年1月至2015年12月每天的价格数据为实例进行分析,并通过训练得到的模型对2016年1月至3月三七价格进行预测,结果表明利用相空间重构方法优化的BP神经网络预测的结果优于传统的ARIMA模型。

In this paper, ARIMA and BP neural networks are used to analyze and forecast the price of Pseudo- ginseng. Firstly, the traditional time series analysis model—autoregressive moving average model is used to predict the price of Pseudo-ginseng, and it is found that it can only extract the linear characteristics of the price curve, and the effect of outlier fitting is poor. BP neural network is applied to solving the nonlinear separable problem. Combining with the fluctuation and non-linearity of the price data of Pseudo-ginseng, the BP neural network is used to analyze and forecast the price of Pseudo-ginseng, and the phase space reconstruction method is used to determine the network input dimension, to determine the network structure better. This paper analyzes the price data from January 2010 to December 2015 in Wenshan area, Yunnan Province, and forecasts the price of January to March in 2016 through the trained model. The results show that the optimized BP neural network by the method of phase space reconstruction is better than the traditional ARIMA model.

文章引用:
刘建中, 李慧超. 基于ARIMA模型与BP神经网络的三七价格预测[J]. 计算机科学与应用, 2017, 7(7): 696-710. https://doi.org/10.12677/CSA.2017.77081

参考文献

[1] 罗群, 游春梅, 官会林. 环境因素对三七生长影响的分析[J]. 中国西部科技, 2010, 9(9): 7-8.
[2] 陈中坚. 三七的无公害栽培[J]. 云南农业科技, 1999(6): 44-45.
[3] 刘峰, 王儒敬, 李传席. ARIMA模型在农产品价格预测中的应用[J]. 计算机工程与应用, 2009, 45(25): 238-239.
[4] 张立杰, 朱新杰. 我国棉花价格长期走势与短期预测——基于差分自回归移动平均模型(ARIMA)的分析[J]. 价格理论与实践, 2012(6): 53-54.
[5] Contreras, J., Espinola, R., Nogales, F.J., et al. (2002) ARIMA Models to Predict Next-Day Electricity Prices. IEEE Power Engineering Review, 18, 1014-1020.
https://doi.org/10.1109/MPER.2002.4312577
[6] Xue, D.-M. and Hua, Z.-Q. (2016) ARIMA Based Time Series Forecasting Model. Recent Advances in Electrical and Electronic Engineering, 9, 93-98.
https://doi.org/10.2174/2352096509999160819164242
[7] Adebiyi, A.A., Adewumi, A.O. and Ayo, C.K. (2014) Comparison of ARIMA and Artificial Neural Networks Models for Stock Price Prediction. Journal of Applied Mathematics, 2014, Article ID: 614342.
[8] SIMON HAYKIN. 神经网络原理[M]. 第2版, 北京: 机械工业出版社, 2006.
[9] 高述涛. CS算法优化BP神经网络的短时交通流量预测[J]. 计算机工程与应用, 2013, 49(9): 106-109.
[10] 李萍, 曾令可, 税安泽, 等. 基于MATLAB的BP神经网络预测系统的设计[J]. 计算机应用与软件, 2008, 25(4): 149-150.
[11] Yang, S.E. and Huang, L. (2005) Financial Crisis Warning Model based on BP Neural Network. Systems Engineering-theory & Practice, 25, 12-19.
[12] Ma, W., Wang, Y. and Dong, N. (2010) Study on Stock Price Prediction Based on BP Neural Network. IEEE International Conference on Emergency Management and Management Sciences, Beijing, 8-10 August 2010, 57-60.
[13] 陈敏, 徐德智, 罗庆云. 时间序列相空间重构及其应用研究[J]. 计算机与信息技术, 2005(11): 11-12.
[14] 郁俊莉, 王其文, 韩文秀. 经济时间序列相空间重构与混沌特性判定研究[J]. 武汉大学学报(理学版), 2004, 50(1): 33-37.
[15] 修春波, 刘向东, 张宇河. 相空间重构延迟时间与嵌入维数的选择[J]. 北京理工大学学报, 2003, 23(2): 219-224.
[16] 徐自励, 王一扬, 周激流. 估计非线性时间序列嵌入延迟时间和延迟时间窗的C-C平均方法[J]. 四川大学学报工程科学版, 2007, 39(1): 151-155.