| 神经网络与时间序列模型在股票预测中的比较 |
| |
| 引用本文： | 王波,张凤玲. 神经网络与时间序列模型在股票预测中的比较[J]. 武汉理工大学学报(信息与管理工程版), 2005, 27(6): 69-73 |
| |
| 作者姓名： | 王波 张凤玲 |
| |
| 作者单位： | 天津大学,理学院,天津,300072 |
| |
| 摘 要： | 首先利用时间序列中的AR IM A模型和人工神经网络建立了两类股票价格预测模型并对一定时期的股票价格进行了预测,然后用4种广为使用的统计评价方法对两类模型的预测性能进行了比较。结果表明,两种模型都取得了很好的整体预测效果,而在趋势预测方面,神经网络则得到了比AR IM A模型更准确的结果。
|
| 关 键 词： | 股票预测 时间序列 模型 神经网络 |
| 文章编号： | 1007-144X（2005）06-0069-05 |
| 收稿时间： | 2005-08-07 |
| 修稿时间： | 2005-08-07 |
Comparison of Artificial Neural Network and Time Series Model for Forecasting Stock Prices |
| |
| Wang Bo,Zhang Fengling. Comparison of Artificial Neural Network and Time Series Model for Forecasting Stock Prices[J]. Journal of Wuhan University of Technology(Information & Management Engineering), 2005, 27(6): 69-73 |
| |
| Authors: | Wang Bo Zhang Fengling |
| |
| Affiliation: | School of Science, Tianjin University, Tianjin 300072, China |
| |
| Abstract: | Two types of price forecasting models based on ARIMA and neural network are developed to predict stock prices in a certain period.Then the forecasting performance of the models is evaluated by using four widely used statistic criterions.As a result,the two models execute close prediction in general sense,but for trend forecasting,the neural network-based model outperforms the ARIMA-based model. |
| |
| Keywords: | stock forecasting time series ARIMA model neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
